Patellar tendon properties distinguish elite from non-elite soccer players and are related to peak horizontal but not vertical power

Patellar tendon properties distinguish elite from non-elite soccer players and are related to... Purpose To investigate potential differences in patellar tendon properties between elite and non-elite soccer players, and to establish whether tendon properties were related to power assessed during unilateral jumps performed in different directions. Methods Elite (n = 16; age 18.1 ± 1.0 years) and non-elite (n = 13; age 22.3 ± 2.7 years) soccer players performed vertical, horizontal-forward and medial unilateral countermovement jumps (CMJs) on a force plate. Patellar tendon (PT) cross- sectional area, elongation, strain, stiffness, and Young’s modulus (measured at the highest common force interval) were assessed with ultrasonography and isokinetic dynamometry. Results Elite demonstrated greater PT elongation (6.83 ± 1.87 vs. 4.92 ± 1.88 mm, P = 0.011) and strain (11.73 ± 3.25 vs. 8.38 ± 3.06%, P = 0.009) than non-elite soccer players. Projectile range and peak horizontal power during horizontal-forward CMJ correlated positively with tendon elongation (r = 0.657 and 0.693, P < 0.001) but inversely with Young’s modulus (r = − 0.376 and − 0.402; P = 0.044 and 0.031). Peak medial power during medial CMJ correlated positively with tendon elongation (r = 0.658, P < 0.001) but inversely with tendon stiffness (r = − 0.368, P = 0.050). Conclusions Not only does a more compliant patellar tendon appear to be an indicator of elite soccer playing status but it may also facilitate unilateral horizontal-forward and medial, but not vertical CMJ performance. These findings should be consid- ered when prescribing talent selection and development protocols related to direction-specific power in elite soccer players. Keywords Patellar tendon · Elongation · Strain · Stiffness · Young’s modulus · Unilateral · Countermovement jump AbbreviationsPT Patellar tendon BFlh Biceps femoris long head PCSA Ph ysiological cross-sectional area CMJ Countermovement jump Peak H-power P eak horizontal power normalised to 2/3 CSA Cross-sectional areabody mass sCSA Cr oss-sectional area normalised to body Peak M-power P eak medial power normalised to body 2/3 2/3 mass mass EMG Electromyography Peak V-power P eak vertical power normalised to body 2/3 mass RMVC R amped maximal voluntary contraction Communicated by Olivier Seynnes. VL Vastus lateralis * Conall F. Murtagh C.F.Murtagh@ljmu.ac.uk Introduction School of Sport and Exercise Sciences, Liverpool John Moores University, Liverpool L3 3AF, UK During the course of a match, an elite soccer player may Liverpool Football Club, Liverpool, UK perform up to ~ 119 maximal accelerations, ~ 35 sprints Department of Rehabilitation Sciences, KU Leuven, (Bradley et al. 2010), ~ 50 forceful changes of direction University of Leuven, 3000 Leuven, Belgium (Withers et  al. 1982), and ~ 16 vertical jumps (Reilly Football Association of Wales, Cardiff, Wales, UK and Thomas 1976). This activity profile implies that the Institute of Sport, Exercise and Health, University College most common type of powerful actions elicited involve London, London, UK Vol.:(0123456789) 1 3 European Journal of Applied Physiology the successive combination of eccentric and concentric performance is inconclusive. Moreover, the methods uti- muscle actions, also known as the stretch-shortening cycle lised in the previous research investigating the association (Cavanagh and Komi 1979). Indeed, elite soccer players between tendon properties and sport-specific performance have been previously shown to outperform non-elite soc- were limited, as they measured the properties of the VL cer players during unilateral jumping activities in different aponeurosis rather than the patellar tendon and, therefore, directions, which induce stretch-shortening cycle actions had to apply a series of corrections to estimate tendon in the muscle–tendon unit of the lower limb (Murtagh elongation (Kubo et al. 2011; Bojsen-Møller et al. 2005; et  al.  2017). The viscoelastic properties of the tendon Stafilidis and Arampatzis 2007). Assessing in vivo human affect the interaction between the contractile and elastic tendon properties directly via ultrasound and isokinetic elements of the muscle–tendon unit, and are thought to dynamometry (Seynnes et al. 2009; Hansen et al. 2006; influence performance during stretch-shortening cycle Wiesinger et al. 2016), on the other hand, would enable activities (Bojsen-Møller et al. 2005). However, the impor- a more precise investigation of a potential relationship tance of tendon properties for determining elite soccer between tendon properties and sport-specific measures of playing status, and their contribution to soccer-associated physical performance. power capabilities, has not yet been investigated. As the vast majority of explosive actions performed dur- A comparison of tendon properties in populations of ing elite soccer match-play involve a single legged push-off well trained and untrained, or different sporting perfor - (Bradley et al. 2010; Withers et al. 1982), and unilateral mance levels, may provide an insight into the importance CMJs in different directions differentiate elite from non- of these characteristics for high-level sports performance elite soccer players (Murtagh et al. 2017), such assessments (Tillin et al. 2010). Within this context, high- (Kubo et al. can be used to assess soccer-associated power (Murtagh 2011) and intermediate-level (Kubo et al. 2000b) sprinters et al. 2017). Indeed, we have previously shown that multi- have a more compliant vastus lateralis (VL) aponeurosis directional unilateral CMJs assess independent power quali- than untrained individuals. More specifically, significant ties (Murtagh et al. 2017) and are underpinned by discrete correlations have been reported between 100 m race per- neuromuscular factors (Murtagh et al. 2018). However, the formance and VL aponeurosis compliance (Kubo et al. importance of tendon properties in elite soccer and their 2000b) and maximal elongation (Stafilidis and Arampatzis contribution to unilateral CMJ performance in different 2007). This suggests that greater elongation of the VL ten- directions is unknown. Such information could inform the don–aponeurosis complex allows the utilisation of greater specific detail of soccer talent identification and develop - energy storage and increases the shortening velocity of the ment protocols relative to the different properties of the knee-extensor muscle–tendon unit during sprint running. PT. Therefore, the aims of this study were (1) to investigate Results from some vertical jump studies are in accordance whether PT properties differed between elite and non-elite with these findings, and report that VL aponeurosis stiff- soccer players; and (2) to establish if relationships existed ness correlated inversely with the calculated difference between the morphological, mechanical, and material prop- in jump height between bilateral vertical countermove- erties of the PT and unilateral jump performance in different ment jumps and squat jumps. These data suggest that the directions in soccer players. greater compliance of the VL tendon–aponeurosis com- plex facilitates the augmentation in jump performance with a countermovement (i.e., when comparing bilateral Methods vertical countermovement jumps vs. squat jumps) (Kubo et al. 1999). In contrast, Bojsen-Møller et al. (2005) dem- Subjects onstrated that VL aponeurosis stiffness correlated posi- tively with bilateral vertical countermovement jump height Twenty-nine male soccer players volunteered to take part in and knee-extensor isometric rate of torque development, this study, which was approved by Liverpool John Moores thus suggesting that a stiffer tendon–aponeurosis complex University Ethics Committee and complied with the Decla- contributes to enhanced muscle output during high force ration of Helsinki. Participants provided written informed isometric and dynamic bilateral vertical jumping tasks consent prior to being assigned to two groups according (Bojsen-Møller et  al. 2005). In agreement, Wiesinger to their level of competition. The elite soccer player group et al. (2016) directly assessed the properties of the patel- (n = 16; age 18.3 ± 1.3  years; height 1.81 ± 0.07  m; body lar tendon in vivo and showed significantly greater tendon mass 76.2 ± 9.7 kg) consisted of players from an English stiffness in elite ski jumping athletes compared to controls. Premier League football academy, who regularly partici- However, Wiesinger et  al. (2016) did not assess sport- pated at U18- and U21-level training and matches. The aver- specific performance directly and the previous research age weekly training/match content for elite soccer players investigating the contribution of tendon properties to jump was: 1 ± 1 soccer matches, 4 ± 1 soccer training sessions, 1 3 European Journal of Applied Physiology 3 ± 1  gym-based resistance training sessions (two upper Methodology body and one lower body), and 4 ± 1 non-resistance train- ing gym sessions (consisting of injury prevention, muscle Countermovement jumps activation and movement training). For more detail on the elite soccer player training frequency/type details, please see On arrival at the laboratory for the second session, all par- our recent study (Brownlee et al. 2018). The average weekly ticipants had their height and body mass measured. Partici- training content for the non-elite soccer players was: 1 ± 1 pants performed three trials of each CMJ type (with 60 s soccer matches, 2 ± 1 soccer training sessions, 1 ± 1 gym- recovery between trials within a single CMJ type, and 180 s based resistance training sessions, and no player reported between jump types), thus performing a total of 18 CMJs participating in any non-resistance training gym ses- (9 unilateral jumps on each leg). The methods for the per- sions. The inclusion criteria for the non-elite soccer player formance and data analysis of unilateral CMJs have been group (n = 13; age 22.4 ± 1.7  years; height 1.74 ± 0.06  m; explained in detail previously (Murtagh et al. 2017). As body mass 72.6 ± 6.6 kg) was to participate in at least 1 h recommended by Jaric et al. (2005), peak power outputs per week of competitive soccer (11-a-side or five-a-side), achieved during CMJs were isometrically scaled to body 0.67 and 1 h per week of soccer-specific or fitness-based training. mass (peak power divided by body mass ). The key per- Non-elite participants were excluded if they did not meet formance variables for the unilateral vertical CMJ were jump these inclusion criteria or had previously played soccer at height [calculated from the impulse–momentum relationship academy, semi-professional, or professional level. All par- derived take off velocity and equation of constant accelera- ticipants had been free of any injury to the lower body within tion methods (Dowling and Vamos 1993)] and peak vertical 0.67 the previous 3 months and had not previously sustained a power isometrically scaled to body mass (peak V-power). serious knee or ankle injury that may be aggravated dur- The key performance variables for horizontal-forward and ing testing procedures, or cause an adverse effect on perfor - medial CMJs were projectile range [calculated using equa- mance. Participants were fully familiarised with all testing tions of constant acceleration (Grimshaw et al. 2004)], peak procedures in a separate session and were asked to complete V-power, and peak horizontal power isometrically scaled to a physical activity and health questionnaire prior to the study body mass [(peak H-power) for horizontal-forward CMJs for screening purposes. This questionnaire allowed us to only] or peak medial power isometrically scaled to body ascertain if each potential participant satisfied the specific mass [(peak M-power) for medial CMJs only]. Projectile inclusion and exclusion criteria. range was used as the criterion performance measure for horizontal-forward and medial CMJs as, unlike when meas- Experimental design uring jump distance using a measuring tape, projectile range is not affected by airborne and landing technique and better All participants attended the laboratory on two occasions, represents the propulsive phase of the jump (Meylan et al. separated by between 3 and 7 days. The r fi st session enabled 2012). the participants to be familiarised with the assessment proto- col, which consisted of three unilateral CMJs in the vertical, Tendon morphology horizontal-forward, and medial directions on each leg, and two knee-extensor ramp maximum voluntary contractions Patellar tendon length, cross-sectional area (CSA), and elon- (RMVCs) on an isokinetic dynamometer. All CMJs were gation measurements were performed using ultrasonography visually demonstrated to the participants by the investiga- (MyLab30, Esaote, Genoa, Italy) with the knee joint set at tor. This session was also used to determine the superior 90° knee flexion (full extension = 0°) and the hip joint set at jumping leg (defined as the limb that produced the high- 85° (supine = 180°). The PT resting length, defined as the est ground reaction force during a unilateral vertical CMJ). distance between the patella apex and the point at which the During the second session, the participants performed all tendon inserts into the tibial tuberosity, was determined by assessments for CMJ and tendon properties on the dominant positioning the 40-mm-wide, 10–15 MHz linear transducer leg, and the results from these assessments were used for in the sagittal plane over the PT and marking the location subsequent analysis. To minimise the influence of previous of the patella apex (0% tendon length) and tibial tuberosity activity, the testing was performed at least 48 h following (100% tendon length) on the skin with a permanent marker any high-intensity exercise. Participants were also instructed pen. Three more locations were then marked on the skin over not to consume alcohol in the 48 h before testing sessions, the tendon (25, 50, and 75% tendon length) to enable PT and to consume no more than 3 mg/kg caffeine (250 mg as CSA to be measured from these axial images using Image an absolute caffeine guideline, e.g., no more than one cup of Analysis Software (ImageJ V.1.45s, National Institute of coffee) in the 5 h prior to the beginning of testing sessions. Health, MD, USA). Transverse scans were taken at each segment until an acceptable scan, in which the borders of the 1 3 European Journal of Applied Physiology whole PT cross section could be clearly identified. This was much from the required linear pattern upon visual inspec- generally achieved between one and three attempts. As the tion (Helland et al. 2013). Torque data during the RMVC stress imposed upon the PT during soccer activity is influ- were sampled using data acquisition software (AcqKnowl- enced by the forces exerted by the quadriceps femoris mus- edge, Biopac Systems Inc., Goleta, CA, USA). Ultrasound cle, in the absence of a published isometric power law for video sequences were recorded at 25 Hz during the RMVC this variable, PT CSA measurements were also isometrically and were synchronised with the RMVC torque data via the scaled to body mass (tendon cross-sectional area divided by administration of a square wave pulse, which was visible 0.67 body mass ; sCSA) (Seynnes et al. 2011). simultaneously on the AcqKnowledge software and the ultrasound monitor (ECG signal). It is acknowledged that Tendon elongation during RMVC this technique may lead to an underestimation of the PT elongation due to unmonitored tibial movements (Onambélé Details and reliability of the measurements have been et al. 2007). Unfortunately, the 4-cm-wide ultrasound trans- documented previously (Seynnes et al. 2009; Hansen et al. ducer did not enable simultaneous scanning of both proximal 2006). In short, the mechanical properties of the tendon were and distal insertions of the PT. However, because the major assessed by measuring the PT elongation during RMVCs. portion of the tendon was scanned in the same way in all par- Prior to performing two RMVCs, a 2-mm-wide strip of sur- ticipants, this technical compromise does not invalidate the gical tape (3M, Neuss, Germany) acted as an echo-absorbent comparative outcome of the data. Furthermore, this method marker and was placed on the skin transversely over the of assessing PT elongation has been shown to have a good tendon as a reference point at ~ 30% tendon length from the test–retest reproducibility by us (Table 1) and others (Reeves proximal end. Immediately before the RMVCs, participants et al. 2003). performed a series of five submaximal isometric contrac - tions to ensure preconditioning of the tendon. The linear Antagonist muscle co-activation transducer was then placed in the sagittal plane over the PT (~ 1 cm of the probe above the patellar apex and ~ 3 cm Antagonist (hamstring) muscle co-activation during RMVC below, incorporating the echo-absorbent marker). Similar to was assessed via electromyography (EMG) activity. The the previous research that prescribed a specific time period biceps femoris long head (BFlh) muscle [representative for ramped isometric contractions when measuring tendon of the knee flexor muscle group (Kellis and Baltzopoulos properties (Hansen et al. 2006; Seynnes et al. 2009; Mal- 1997)] was identified via palpation during a submaximal liaras et al. 2013), the RMVCs lasted 6 s in total with 2 min knee flexion with the participant in the prone position. rest between each contraction. Loading rates (Nm/s) were, After preparing the skin surface (shaving, lightly abrad- therefore, dependent on the participant’s maximal volun- ing, and cleansing with 70% ethanol) at 2/3 muscle length tary force capacity. Visual feedback of force production was from the proximal end [SENIAM guidelines (Freriks et al. displayed on a screen in front of the participants to ensure 1999)], two bipolar Ag–AgCl surface electrodes (Neu- that all RMVCs were performed at a constant loading rate. roline, Medoicotest, Rugmarken, Denmark) were placed Trials were discarded when the torque trace deviated too 20  mm apart along the sagittal axis of the muscle belly Table 1 Test–retest Tendon property Typical error (95% CI) CV (%) ICC (95% CI) reproducibility of patellar tendon measurements in eight CSA (25% TL) 1.460 (0.966–2.972) mm 1.234 0.997 (0.987–0.999) healthy, recreationally active CSA (50% TL) 1.685 (1.114–3.429) mm 1.694 0.978 (0.898–0.995) young men CSA (75% TL) 2.591 (1.713–5.273) mm 1.996 0.988 (0.941–0.998) Mean CSA 1.190 (0.787–2.421) mm 1.004 0.996 (0.980–0.999) Stress 0.4065 (0.269–0.827) MPa 0.968 0.996 (0.982–0.999) Elongation 0.322 (0.213–0.657) mm 5.653 0.986 (0.935–0.997) Strain 0.704 (0.465–1.432) % 5.506 0.986 (0.936–0.997) Stiffness 63.6 (39.7–156.1) N/mm 4.477 0.988 (0.912–0.998) Young’s modulus 0.032 (0.021–0.071) GPa 5.930 0.979 (0.889–0.996) Measurements were taken on two separate occasions (at the same time of day) within a 12-day period. CSA measurements were taken at rest with the knee flexed at 90° (0° = full extension). Tendon stress, elonga- tion, strain, stiffness, and Young’s modulus were all calculated using the highest common force (interval) CV coefficient of variation, ICC intra-class correlation coefficient, CSA cross-sectional area, TL tendon length, MPa megapascals, GPa gigapascals P < 0.001 for all ICCs 1 3 European Journal of Applied Physiology and a reference electrode placed on the lateral tibial con- However, we do acknowledge that this technique is limited, dyle. The root mean square (RMS) of the BFlh EMG signal as we are assuming that the force elongation relationship was recorded at each 10% knee extension R MVC and for between 2599–3248  N is linear (Pearson and Onambélé 500 ms around the peak torque during a knee flexion MVC 2012). PT elongation was calculated as the change in tendon at 90° knee flexion. Assuming a linear relationship between length at (3248 N) from resting length. Tendon strain was EMG and torque output (Kellis and Baltzopoulos 1997), calculated as the change in tendon length at the highest com- the torque generated by the antagonists during each 10% mon RMVC force level relative to the original tendon length knee extension RMVC was estimated by dividing the BFlh (∆L/L ), and expressed as a percentage. Young’s modulus EMG during knee-extensor RMVC by the BFlh maximal (E) was calculated by multiplying stiffness (k ) with the ratio EMG (BFlh EMG during knee flexion MVC) and multiply - of the resting tendon length (l ) to mean tendon CSA, i.e., ing this ratio by the knee flexion MVC torque at each 10% E = k × (l /s). RMVC. This torque was added to the knee-extensor torque value at the relevant 10% RMVC to provide the gross knee- Test–retest reproducibility extensor RMVC torque. Dividing the RMVC torque by the PT moment arm at 90° knee flexion [assumed to be 0.048 m To establish the test–retest reproducibility of our methods based on MRI measurements in young healthy men (Erskine for measuring the morphological, mechanical, and material et al. 2009)] provided the PT force at each 10% RMVC. Due properties of the PT, we recruited an additional eight rec- to technical issues with some EMG recordings, EMG data reationally active, young, healthy men (age 28.0 ± 3.0 years; were only available in a subsample of elite (n = 12) and non- body mass 77.1 ± 7.4 kg; height 1.79 ± 0.04 m). All par- elite (n = 8) participants. As a limited sample size will nega- ticipants attended the laboratory on two separate occasions, tively impact the statistical power of correlation analyses at the same time of day, in the same standardised condi- (Guadagnoli and Velicer 1988), and investigating the rela- tions as specified for the elite and non-elite soccer players, tionship between tendon properties and power during CMJs with a period of 7–12 days in between each testing session. was one of the main aims of this study, we have reported Measurements were performed by the same researcher, as tendon properties without accounting for antagonist muscle described in the “Tendon morphology”, “Tendon elonga- co-activation [which is not uncommon practice (Stenroth tion during RMVC”, and “Analysis of tendon data” sections. et al. 2012, 2015; Peltonen et al. 2012)] in an effort to max- Inter-day reliability for each measurement was expressed as imise statistical power. However, to support these findings, typical error (TE), coefficient of variation (CV), and intra- we have also reported the results from correlation analyses class correlation coefficient (ICC, model: two-way mixed; in the subsample, where antagonist muscle co-activation type: absolute agreement) with 95% confidence intervals was accounted for. (CIs). These data are presented in Table  1, and for all variables, the TE and CV were low, and the ICC was high Analysis of tendon data (≥ 0.979) with narrow CIs, thus indicating that our methods were highly reproducible. PT elongation was recorded as the distance the patella apex moved from the external marker every 10% RMVC. Statistical analyses The video frame that corresponded to each 10% RMVC was exported as a portable network graphics (.png) file and The mean and standard deviation (s) were calculated for all the distance from the patella apex to the external marker variables in both the whole cohort, and a subsample where was measured using image analysis software (ImageJ v. antagonist co-activation was accounted for. All data were 1.47, National Institute of Health, MD, USA). Individual tested for normality using the Shapiro–Wilks normality test. force–elongation curves were fitted with a second-order For tendon CSA, tendon sCSA, and tendon elongation at polynomial (R > 0.95 in all cases). Tendon stiffness each 10% RMVC interval, a two-way mixed ANOVA was (∆Force/∆PT length) values were obtained over the high- used to determine a main effect of athlete status (between est 20% common force interval in the weakest participant factor: elite vs. non-elite soccer players), tendon location with the lowest maximal RMVC peak force (2599–3248 N). [within factor: CSA at 0, 25, 50, 75, and 100% tendon length This approach, as performed by others (e.g., Seynnes et al. (this includes mean tendon CSA comparison)] or tendon 2009), was necessary to avoid having to extrapolate some elongation (within factor: tendon elongation at 10, 20, 30, data points beyond the visible part of the force–elonga- 40, 50, 60, 70, 80, 90, and 100% RMVC), and an athlete tion curve relationship (which would have been required status × tendon location, or athlete status × tendon elonga- for the weaker participants). It also allowed us to compare tion at a specific % of RMVC, interaction. If a significant the mechanical and material properties of elite and non- interaction occurred, simple main effects and pairwise elite player tendons under the same forces and conditions. comparisons with Bonferroni adjustment were performed 1 3 European Journal of Applied Physiology to reveal differences. If a significant main effect existed for tendon CSA location or tendon elongation at a specific % of RMVC, Bonferroni post hoc tests were used to establish the difference in CSA or tendon elongation between tendon locations or % RMVC force levels, respectively. For tendon force, length, stiffness, elongation, strain, and Young’s mod- ulus, independent t tests were used to determine differences between elite and non-elite soccer players, in both the whole cohort and subsample (where antagonist co-activation was measured). Pearson’s correlations were used to determine relationships between jump performance variables (height or projectile range, peak V-power, peak H-power, or peak M-power) and tendon mean CSA and CSA at 0, 25, 50, 75, Fig. 1 Patellar tendon cross-sectional area (CSA) along its length at 25% intervals in elite (n = 16, circles, mean + SD) and non-elite and 100% tendon length, strain, stiffness, and Young’s mod- (n = 13, open squares, mean − SD) players. *Main effect, elite players ulus. Statistical analysis was completed using SPSS version significantly greater CSA than non-elite players (P < 0.05) 21 (SPSS Inc., Chicago, IL, USA) and the significance level was set at P ≤ 0.05. Table 2 Patellar tendon morphological, mechanical, and material properties calculated in the whole cohort [elite (n = 16) and non-elite (n = 13)]; mean ± SD Results Tendon property Elite Non-elite Anthropometry Mean CSA (mm ) 115 ± 6* 112 ± 3 2 0.67 Mean sCSA (mm /kg ) 6.35 ± 0.44 6.41 ± 0.02 Elite soccer players were significantly taller (P = 0.009) Resting tendon length (mm) 58.5 ± 7.1 58.2 ± 4.4 and had significantly longer femur length (46.8 ± 2.1 vs. Stiffness (N/mm) 1269 ± 607 1707 ± 792 44.0 ± 1.7  cm; P = 0.001) than non-elite. However, there Elongation (mm) 6.83 ± 1.87* 4.92 ± 1.88 was no difference in body mass between elite and non-elite Strain (%) 11.73 ± 3.25** 8.38 ± 3.06 soccer players (P = 0.267). Stress (MPa) 28.29 ± 1.34* 29.15 ± 0.72 Young’s modulus (GPa) 0.64 ± 0.31 0.88 ± 0.39 Differences in patellar tendon properties MVC tendon force (N) 5728 ± 1522 4800 ± 989 between elite and non‑elite players CSA cross-sectional area, sCSA cross-sectional area isometrically 0.67 scaled to body mass , MVC maximal voluntary contraction Tendon morphology *A significant difference when comparing elite and non-elite (P < 0.05) For tendon CSA, there was a main effect of athlete sta- **A significant difference between elite and non-elite (P < 0.01) tus, with elite soccer players demonstrating ~ 2.6% greater mean CSA (F = 4.439, P = 0.045; Fig. 1; Table 2). There 1,27 was also a main effect of tendon location ( F = 105.36, 0 vs. 75%, P < 0.001; 0 vs. 100%, P < 0.001; 25 vs. 75%, 4,108 P < 0.001) with pairwise post hoc analyses, revealing that P < 0.001; 25 vs. 100%, P < 0.001; 50 vs. 75%, P < 0.001; tendon CSA varied along the length of the tendon (0 vs. 50 vs. 100%, P < 0.001; 75 vs. 100%, P = 0.031); except at 25%, P = 0.001; 0 vs. 50%, P < 0.001; 0 vs. 75%, P < 0.001; 25 vs. 50% tendon length where there was a non-significant 0 vs. 100%, P < 0.001; 25 vs. 75%, P < 0.001; 25 vs. 100%, difference (P = 0.185). However, there was no significant 0.67 P < 0.001; 50 vs. 75%, P < 0.001; 50 vs. 100%, P < 0.001; main effect of athlete status (elite 0.063 ± 0.004 mm/kg ; 0.67 75 vs. 100%, P = 0.007; Fig. 1); except at 25 vs. 50% ten- non-elite 0.063 ± 0.004 mm/kg , F = 0.006, P = 0.941) 1,27 don length where there was a non-significant difference and no significant interaction between athlete status and ten- (P = 0.063; Fig.  1). There was no significant interaction don CSA location (F = 0.720, P = 0.580). Furthermore, 4,108 between tendon CSA and athlete status (F = 0.720, tendon resting length relative to femur length did not differ 4,96 P = 0.565). Tendon resting length did not differ between between groups (t = − 1.520; P = 0.140). groups (t = 0.119; P = 0.906; Table 2). For tendon sCSA, there was a main effect of tendon loca - Tendon mechanical and material properties tion (F = 106.07, P < 0.001) with pairwise post hoc 4,108 analyses, revealing that tendon CSA varied along the length We present differences in the tendon mechanical and mate- of the tendon (0 vs. 25%, P = 0.024; 0 vs. 50%, P < 0.001; rial properties between elite and non-elite soccer players 1 3 European Journal of Applied Physiology calculated in both the whole cohort (elite: n = 16; non-elite: Relationships between patellar tendon properties n = 13) and subsample (elite: n = 12; non-elite, n = 8) where and direction‑specific jump performance antagonist co-activation was accounted for. Elite soccer players demonstrated significantly greater tendon elonga - The positive and inverse relationships between jump per- tion and strain than non-elite players (whole cohort: elonga- formance variables and tendon properties from the whole tion, P = 0.011; strain, P = 0.009; Table 2; subsample: elon- cohort are displayed in Table 3. Additional gu fi res have been gation, 6.72 ± 1.94 vs. 4.76 ± 1.97  mm, P = 0.041; strain, included to illustrate the spread of data for specific relation - 11.38 ± 3.27 vs. 8.03 ± 3.25%, P = 0.037). Moreover, when ships. In the subsample, correlations between jump perfor- we compared tendon elongation at each 10% RMVC inter- mance variables and tendon properties are also illustrated val in the whole cohort, there was a significant main effect in the text. of athlete status (F = 93.00, P < 0.001, Fig. 2) and tendon In the whole cohort and subsample, unilateral horizontal- 9,243 elongation (F = 11.13, P < 0.001, Fig. 2), but no significant forward jump performance variables correlated positively 9,27 interaction effect between tendon elongation and athlete sta- with tendon elongation (whole cohort: Fig.  3a; Table  3; tus (F = 1.63, P = 0.108). This finding suggests that elite subsample: projectile range, r = 0.733, P < 0.001, peak 9,243 soccer player tendon elongation was greater than non-elite H-power: r = 0.734, P < 0.001) and strain (whole cohort: soccer players at all 10% levels of RMVC. However, there Table 3; subsample: projectile range: r = 0.765, P < 0.001, were no differences between elite and non-elite soccer players peak H-power: r = 0.677, P = 0.001), but inversely with regarding tendon stiffness (whole cohort: P = 0.104; Table 2; tendon stiffness (whole cohort: Fig.  3b; Table 3; subsam- subsample: P = 0.496). There was a non-significant tendency ple: projectile range: r = − 0.576, P = 0.008, peak H-power: for Young’s modulus (P = 0.075; Table 2) to be lower and r = − 0.579, P = 0.007) and Young’s modulus (whole cohort: maximum tendon force (P = 0.069; Table 2) to be higher in Fig. 3c; Table 3; subsample: projectile range: r = − 0.559, elite compared to non-elite soccer players in the whole cohort. P = 0.010, peak H-power: r = − 0.490, P = 0.028). Simi- In the subsample, elite soccer players produced significantly larly, unilateral medial CMJ performance variables cor- greater maximum tendon force in comparison to non-elites related positively with tendon elongation (whole cohort: (6374 ± 1658 vs. 4817 ± 607 N, P = 0.021). Fig.  4a; Table  3; subsample: projectile range: r = 0.513, For a comparison of unilateral jump performance between P = 0.021, peak M-power: r = 0.633, P = 0.003) and strain elite and non-elite soccer players, please refer to our data (whole cohort: Fig. 4b; Table 3; subsample: peak M-power: presented in a previously published study specifically aimed r = 0.570, P = 0.009), but inversely with tendon stiffness to determine whether jump performance could distinguish (whole cohort: Fig. 4c; Table 3; subsample: tendency for between soccer playing levels (Murtagh et al. 2017). a significant correlation with peak M-power: r = − 0.432, P = 0.057). Mean tendon CSA, CSA at 75% and CSA at 100% correlated positively with unilateral horizontal-for- ward CMJ peak V-power and unilateral medial CMJ peak V-power (Table 3) in the whole cohort. Furthermore, tendon CSA at 50 and 100% correlated positive with unilateral ver- tical jump peak V-power (Table 3). There were no significant correlations between any jump variable and tendon CSA at 0 or 25% (r ≤ 0.307; P ≥ 0.116) (Table 3). Discussion The aims of our study were to compare PT properties between elite and non-elite soccer players, and to establish if these properties were related to unilateral jump perfor- mance in different directions. Our results showed that ten- don CSA, elongation and strain were greater in elite soccer 0 1 2 3 4 5 6 7 8 9 10 11 players. These specific tendon properties may, therefore, Elongation (mm) be considered determinants of elite soccer playing status. Tendon elongation and strain also correlated positively with Fig. 2 In vivo patellar tendon force–elongation relation in elite peak horizontal power and medial power during unilateral (n = 16; circles) and non-elite (n = 13; open squares) soccer play- horizontal and unilateral medial CMJ, respectively, while ers. Significant main effect for athlete status (P < 0.05). Data are mean + SD tendon stiffness and Young’s modulus correlated inversely 1 3 Tendon force (N) European Journal of Applied Physiology Table 3 Correlations between unilateral countermovement jump (CMJ) performance measures and patellar tendon properties in elite (n = 16) and non-elite (n = 13) soccer players Jump type Unilateral Vertical CMJUnilateral Horizontal CMJUnilateral Medial CMJ Tendon property Jump height Peak V-power Projectile range Peak H-power Peak V-power Projectile range Peak M-power Peak V-power (cm) (cm) (W/kg) (W/kg) (W/kg) (W/kg) (W/kg) (cm) Stiffness, N/mm r = -0.019 r = 0.168 r = -0.333 r = -0.409 r = 0.013 r = -0.150 r = -0.368 r = 0.207 P = 0.924 P = 0.393 P = 0.077 P = 0.028 P = 0.948 P = 0.436 P = 0.050 P = 0.282 Elongation, mm r = 0.114 r = -0.021 r = 0.657 r = 0.693 r = 0.067 r = 0.496 r = 0.658 r = 0.078 P = 0.473 P = 0.917 P < 0.001 P < 0.001 P = 0.731 P = 0.006 P < 0.001 P = 0.687 Strain, % r = 0.248 r = 0.066 r = 0.698 r = 0.674 r = 0.180 r = 0.409 r = 0.616 r = 0.097 P = 0.204 P = 0.740 P < 0.001 P < 0.001 P = 0.350 P = 0.027 P < 0.001 P = 0.615 Young’s modulus, GPa r = -.081 r = 0.109 r = -0.376 r = -0.402 r = -0.081 r = 0.089 r = -0.344 r = 0.195 P = 0.681 P = 0.583 P = 0.044 P = 0.031 P = 0.675 P = 0.648 P = 0.068 P = 0.311 Mean CSA, mm r = 0.165 r = 0.339 r = 0.109 r = -0.038 r = 0.381 r = -0.009 r = -0.020 r = 0.401 P = 0.402 P = 0.076 P = 0.078 P = 0.845 P = 0.040 P = 0.961 P = 918 P = 0.031 CSA at 50% tendon length r = 0.233 r = 0.376 r = 0.066 r = 0.006 r = 0.213 r = 0.120 r = 0.067 r = 0.308 mm P = 0.233 P = 0.049 P = 0.733 P = 0.975 P = 0.267 P = 0.534 P = 0.732 P = 0.104 CSA at 75% tendon length r = 0.024 r = 0.082 r = 0.251 r = 0.026 r = 0.396 r = -0.048 r = -0.036 r = 0.416 mm P = 0.904 P = 0.677 P = 0.190 P = 0.892 P = 0.033 P = 0.804 P = 854 P = 0.025 CSA at 100% tendon length r = 0.189 r = 0.414 r = 0.150 r = -0.013 r = 0.496 r = -0.040 r = -0.172 r = 0.654 mm P = 0.334 P = 0.029 P = 0.437 P = 0.948 P = 0.006 P = 0.837 P = 0.372 P < 0.001 Significant correlations are highlighted in grey. Significant inverse correlations are highlighted in bold 0.67 CSA cross-sectional area, peak V-power peak vertical power isometrically scaled to body mass , peak H-power peak horizontal power iso- 0.67 0.67 metrically scaled to body mass , peak M-power peak medial power isometrically scaled to body mass with horizontal power during unilateral horizontal CMJ. and tendon size (see Wiesinger et al. 2015 for review), thus Only PT CSA at 50% and 100% tendon length was related potentially suggesting that long-term exposure to elite soccer to unilateral vertical CMJ performance. Our study suggests training results in greater PT tendon CSA. However, when 0.67 that a more compliant PT facilitates greater unilateral CMJ tendon CSA was normalised to body mass , there was no performance in the horizontal-forward and medial direc- significant difference in tendon CSA between cohorts at any tions, but PT stiffness does not contribute to vertical CMJ tendon location (no significant main effect of athlete status performance. and no significant interaction between athlete status and ten- Comparing the physiological capabilities in soccer play- don CSA). Although we found no significant difference in 0.67 ers competing at different performance levels may provide body mass between the elite and non-elite players, this an insight into the physiological factors that determine approach was necessary due to the close relationship found elite soccer playing status. Tendon CSA was significantly previously between body mass and tendon CSA (Rosager greater in elite compared to non-elite soccer players at all et al. 2002). Elite players were significantly taller than non- tendon locations (main effect of athlete status and no sig- elites and the difference in tendon CSA between elite and nificant interaction between athlete status and tendon CSA). non-elite players may, therefore, be related to individual Although the contents of elite and non-elite player habitual differences in body size rather than adaptations to chronic weekly training schedules seem similar (please refer to the habitual loading patterns. Interestingly, in agreement with “Methods” section), elite soccer players perform a higher other literature (Wang et al. 2013), although elite players volume of soccer training sessions and, unlike non-elite displayed greater stature and femur length compared to non- players, also partake in non-resistance-based gym training. elites, there was no difference in resting tendon length, thus Elite players are, therefore, exposed to a greater number of potentially suggesting that the length of the PT in soccer cyclic lower limb loading cycles and high impact explosive players is not related to their body dimensions. Neverthe- actions per week, both of which have been shown to poten- less, without a control group (who did not partake in soccer- tially stimulate increments in tendon CSA (Magnusson and specific training) to compare with, or longitudinal training Kjaer 2003; Couppe et al. 2008; Wiesinger et al. 2016). Sub- study, it still remains unknown if long-term soccer training sequently, our results may agree with the previous literature stimulates PT CSA adaptation. documenting the relationship between habitual daily loading 1 3 European Journal of Applied Physiology Fig. 4 Relationship between unilateral medial (UM) counter- movement jump (CMJ) peak M-power and tendon elongation (a, Fig. 3 Relationship between unilateral horizontal (UH) countermove- r = 0.658, P < 0.001), strain (b, r = 0.616, P < 0.001), and stiffness (c, ment jump (CMJ) peak H-power and tendon elongation (a, r = 0.693, r = − 0.368, P = 0.050) in elite (n = 16, circles) and non-elite (n = 13, P < 0.001), stiffness (b, r = − 0.409, P = 0.028), and Young’s Modu- open squares) players. Peak M-power peak medial power isometri- lus (c, r = − 0.402, P = 0.031) in elite (n = 16, circles) and non-elite 0.67 cally scaled to body mass (n = 13, open squares) players. Peak H-power peak horizontal power 0.67 isometrically scaled to body mass Despite greater tendon CSA, tendon strain was higher in of the common highest 20% force interval of the weakest elite compared to non-elite soccer players. This illustrates participant: 2599–3248 N) were applied to the PT, there that elite soccer players have PT structures that elongate was no significant difference in the change in elongation more, relative to their original length, compared to non- between elite and non-elite players. It, therefore, appears elites at a given tendon force value (highest common ten- that the greater magnitude of difference in PT elongation don force = 3248 N). Although elite soccer players displayed between elite and non-elite players occurs at lower force greater PT elongation, we found no significant difference levels (below the common force interval of the weakest par- in tendon stiffness between elite and non-elite soccer play - ticipant). Young’s modulus (i.e., the relation between stress ers. This suggests that when specific forces (representative and strain), which represents the material properties of the 1 3 European Journal of Applied Physiology tendon independent of its dimensions, provides a more accu- unilateral horizontal-forward CMJ peak H-power and uni- rate representation of the in vivo function of the PT when lateral medial CMJ peak M-power, while tendon stiffness comparing elite and non-elite soccer players (Foster et al. was inversely related to unilateral horizontal-forward CMJ 2014). Our findings showed that there was a tendency for peak H-power and unilateral medial CMJ peak M-power. elite soccer players to display lower Young’s modulus than These findings suggest that more compliant PTs facilitate non-elite players. The greater elongation and strain dem- greater unilateral horizontal and unilateral medial CMJ onstrated by elite in comparison to non-elite soccer players performance in soccer players. Moreover, Young’s modu- may, therefore, be due to differences in the microstructure lus was inversely related with unilateral horizontal CMJ of the tendon, including less collagen cross linking and projectile range and peak H-power, thus suggesting that less fibril packing (Reed and Iozzo 2002). Moreover, as the a tendon with more elastic properties enhances unilateral major difference in the change in PT elongation between horizontal CMJ performance. elite and non-elite players occurs at lower force levels, it may As acceleration and sprinting activities require the pro- be that elite soccer players display PTs with more collagen duction of high levels of horizontal-forward power (Buch- crimping (Kastelic et al. 1980). However, given the non- heit et al. 2014), these findings are in accordance with the significant difference in Young’s modulus between groups, previous studies reporting that 100 m sprint performance is this theory remains speculative. positively related to VL aponeurosis compliance (Kubo et al. Considering the sprint demands of elite soccer match- 2000b) and maximal elongation (Stafilidis and Arampatzis play (Bradley et  al. 2010), the greater PT elongation in 2007). Real-time ultrasonography observations of tendon elite compared to non-elite soccer players in our study is behaviour in  vivo has shown that, during high-intensity consistent with previous studies, reporting that the VL ten- jumping movements whereby the range of joint motion is don–aponeurosis complex of elite sprinters elongates more small, the VL muscle fascicles lengthen only marginally, if than both non-elite sprinters (Stafilidis and Arampatzis at all during the eccentric phase (Finni et al. 2003), and are 2007) and untrained participants (Kubo et al. 2011). While thought to function quasi-isometrically. Horizontal-forward it has been well documented that resistance and plyometric CMJs have been shown to require ~ 10° less knee flexion training interventions increase PT stiffness (see Wiesinger than vertical CMJs (Fukashiro et al. 2005). Hence, unilateral et  al. 2015 for review), no specific training intervention horizontal-forward and possibly unilateral medial CMJs may has been shown to increase the elongation properties of require a quasi-isometric contraction of the knee-extensor the PT structures (Seynnes et al. 2009; Reeves et al. 2003), muscle group, inducing greater tendon lengthening (Reeves and only bed rest (Reeves et al. 2005), detraining (Kubo and Narici 2003), thus allowing the tendon to store more et al. 2010), and ageing (Reeves et al. 2003) are known to potential energy and recoil at greater speeds. This would induce these changes. Therefore, as the CC genotype of the enable the tendon to act as a power amplie fi r during horizon - COL5A1 (rs12722) genetic variant (which encodes the pro tal-forward and medial CMJs (Nagano et al. 2004). Moreo- alpha chain of the type V collagen) has previously associated ver, it has also been reported that during the initial concen- with greater PT elongation and lower stiffness (Kubo et al. tric phase of stretch-shortening cycle exercises, the rapid 2013), it is possible that elite soccer players have a genetic shortening of the tendon contributes to lowering the short- predisposition to tendon structures with greater elongation ening velocity of the muscle fibres to near isometric condi- properties compared to non-elite soccer players. However, tions (Kawakami et al. 2002). Therefore, a more compliant the research associating specific genetic variants and PT PT will have a capacity for greater elongation, allowing the properties remains inconclusive (Foster et al. 2014) and it knee-extensor muscle fibres more time to develop greater cannot be discounted that certain types or combinations of forces during the concentric propulsive phase of sprinting long-term soccer development training stimuli, which have (Kubo et al. 2011), unilateral horizontal-forward CMJ and not been investigated to date, could induce such tendon unilateral medial CMJ activities. Hence, our data support adaptations. Nevertheless, our findings highlight the poten- a notion that a more compliant PT could enhance soccer tial importance of assessing PT elongation in elite soccer player performance levels during activities that depend heav- talent selection and development protocols. ily upon horizontal push-off capacity. Unilateral CMJ tasks oriented in different directions While tendon CSA was not related to unilateral hori- have been suggested to represent a measurement of soccer- zontal-forward or medial CMJ peak H-power or M-power, associated power (Murtagh et al. 2017), and the relation- respectively, distal tendon CSA was positively associated ship between tendon properties and unilateral CMJ perfor- with the ability to produce peak vertical power during uni- mance in different directions are, therefore, of interest for lateral vertical, horizontal and medial CMJs (Table 3). How- the assessment and development of muscular power in elite ever, there was no relationship between any mechanical or soccer players. Our study showed that in soccer players, material PT properties with peak V-power during any CMJ. PT elongation and strain were both positively related to We have previously shown that unilateral vertical and medial 1 3 European Journal of Applied Physiology CMJ peak V-power is positively related to quadriceps femo- that long-term elite soccer training stimulates such adapta- ris muscle size (Murtagh et al. 2018). Considering the rela- tions and may support the hypothesis of functionally driven tionship between body size and tendon CSA (Rosager et al. tendon adaptation proposed by Wiesinger et  al. (2016). 2002), players with larger quadriceps may also have greater However, future research is needed to confirm this theory PT CSA. Hence, the ability to produce peak V-power may as when we accounted for body size, no significant differ - be more closely related to the neuromuscular properties of ence between elite and non-elite players was found. We also the knee extensors rather than the properties of the PT and show that a more compliant PT appears to facilitate horizon- this may be the reason why we found a positive relationship tal-forward and medial jump performance, and therefore, between tendon CSA and peak V-power. The lack of direct practitioners aiming to develop unilateral explosive perfor- relationships between PT mechanical and material proper- mance in these directions should, perhaps, reconsider the ties and unilateral vertical CMJ performance in the soccer prescription of training interventions that may induce a stiff- players in our study is in accordance with some (Kubo et al. ening of the PT. We recommend that future research should 2000a), but not all the previous studies (Bojsen-Møller et al. also investigate training interventions that may increase PT 2005). Bojsen-Møller et al. (2005) found that a stiffer VL elasticity. tendon–aponeurosis complex contributed to enhanced peak force and power outputs during high-intensity isometric and dynamic bilateral vertical jumping tasks, respectively. Conclusion Discrepancies between the findings of Bojsen-Møller et al. (2005) and our results are likely due to methodological dis- We have shown that patellar tendon CSA, elongation, and parities. Whilst we analysed ultrasound images of the PT, strain are greater in elite compared to non-elite soccer play- Bojsen-Møller et al. (2005) measured the VL aponeurosis ers, and are, therefore, important indicators of U18 and U21 and, therefore, approximated tendon elongation. Conse- elite soccer playing status. More compliant patellar tendon quently, it is possible that their stiffness measurements do properties were positively related to performance in jumps not account for total displacement of the tendon and may with a horizontal push-off, yet they were not related to ver - underestimate tendon compliance. tical jump performance. We show that a more compliant We do acknowledge some limitations to our study that patellar tendon will likely benefit elite soccer players across could inform future research. Our investigation only includes all soccer-specific activities involving a powerful single-leg elite players from one soccer club and should be replicated push off in the horizontal direction. in a wider range of soccer institutions. Moreover, we have Acknowledgements The authors wish to thank Raja Azidin and Chris not directly investigated the capacity of the tendon to store Nulty for their help and expertise during the lab testing procedures, and release energy in vivo and future research should also Remy Tang and Neil Critchley for their co-operation with the recruit- measure tendon hysteresis [as performed by Wiesinger ment of elite players, and the participants from Liverpool Football Club et al. (2017)]. The loading rates during the ramped isomet- Academy and Liverpool John Moores University. ric voluntary contractions were dependent on the partici- pants maximal voluntary force level. The previous literature Compliance with ethical standards shows that there are differences in the stiffness and Young’s Conflict of interest The authors declare that they have no conflict of modulus calculations when ramped isometric contractions interest. are performed at 50 N m/s compared to 80 and 110 N m/s (Kösters et al. 2014). Subsequently, we cannot discount that Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the insti- the different loading rates may have affected calculations tutional and/or national research committee and with the 1964 Helsinki of stiffness and Young’s modulus. However, as the loading declaration and its later amendments or comparable ethical standards. rates between participants in our study (29.7–71.3 N m/s) are within a narrow range, we believe it is unlikely that this Informed consent Written informed consent was obtained from all individual participants included in the study. method of assessment compromised the validity of our data. Moreover, as the previous studies adopted similar method- ologies (Hansen et al. 2006; Seynnes et al. 2009; Malliaras Open Access This article is distributed under the terms of the Crea- et al. 2013), we believe that our data are comparable to cur- tive Commons Attribution 4.0 International License (http://creat iveco mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- rent research studies. tion, and reproduction in any medium, provided you give appropriate Our study suggests that novel soccer talent selection pro- credit to the original author(s) and the source, provide a link to the tocols should include a measurement of PT properties, and Creative Commons license, and indicate if changes were made. elite soccer clubs should aim to recruit, and develop, play- ers with more elastic/compliant PTs. As elite players dis- played greater absolute tendon CSA, our research suggests 1 3 European Journal of Applied Physiology exercise in humans reveals a significant role for tendon elastic- References ity. J Physiol 540(2):635–646 Kellis E, Baltzopoulos V (1997) The effects of antagonist moment Bojsen-Møller J, Magnusson SP, Rasmussen LR, Kjaer M, Aagaard on the resultant knee joint moment during isokinetic test- P (2005) Muscle performance during maximal isometric and ing of the knee extensors. Eur J Appl Physiol Occup Physiol dynamic contractions is influenced by the stiffness of the tendi- 76(3):253–259 nous structures. J Appl Physiol 99(3):986–994 Kösters A, Wiesinger H, Bojsen-Møller J, Müller E, Seynnes OR Bradley PS, Di Mascio M, Peart D, Olsen P, Sheldon B (2010) High- (2014) Influence of loading rate on patellar tendon mechanical intensity activity profiles of elite soccer players at different per - properties in vivo. Clin Biomech 29(3):323–329 formance levels. J Strength Cond Res 24(9):2343–2351. https :// Kubo K, Kawakami Y, Fukunaga T (1999) Influence of elastic proper - doi.org/10.1519/JSC.0b013 e3181 aeb1b 3 ties of tendon structures on jump performance in humans. J Appl Brownlee TE, Murtagh CF, Naughton RJ, Whitworth-Turner CM, Physiol 87(6):2090–2096 O’Boyle A, Morgans R, Morton JP, Erskine RM, Drust B (2018) Kubo K, Kanehisa H, Kawakami Y, Fukunaga T (2000a) Elastic prop- Isometric maximal voluntary force evaluated using an isomet- erties of muscle–tendon complex in long-distance runners. Eur J ric mid-thigh pull differentiates English Premier League youth Appl Physiol 81(3):181–187 soccer players from a maturity-matched control group. Sci Med Kubo K, Kanehisa H, Kawakami Y, Fukunaga T (2000b) Elasticity Football:1–7 of tendon structures of the lower limbs in sprinters. Acta Physiol Buchheit M, Samozino P, Glynn JA, Michael BS, Al Haddad H, Men- Scand 168(2):327–335 dez-Villanueva A, Morin JB (2014) Mechanical determinants of Kubo K, Ikebukuro T, Yata H, Tsunoda N, Kanehisa H (2010) Time acceleration and maximal sprinting speed in highly trained young course of changes in muscle and tendon properties during strength soccer players. J Sports Sci 32(20):1906–1913 training and detraining. J Strength Cond Res 24(2):322–331 Cavanagh P, Komi P (1979) Electromechanical delay in human skeletal Kubo K, Ikebukuro T, Yata H, Tomita M, Okada M (2011) Morpho- muscle under concentric and eccentric contractions. Eur J Appl logical and mechanical properties of muscle and tendon in highly Physiol Occup Physiol 42(3):159–163 trained sprinters. J Appl Biomech 27(4):336–344 Couppe C, Kongsgaard M, Aagaard P, Hansen P, Bojsen-Moller J, Kubo K, Yata H, Tsunoda N (2013) Effect of gene polymorphisms on Kjaer M, Magnusson SP (2008) Habitual loading results in tendon the mechanical properties of human tendon structures. Springer- hypertrophy and increased stiffness of the human patellar tendon. plus 25(2):343 J Appl Physiol 105(3):805–810 Magnusson SP, Kjaer M (2003) Region-specific differences in Achil- Dowling JJ, Vamos L (1993) Identification of kinetic and temporal les tendon cross-sectional area in runners and non-runners. Eur J factors related to vertical jump performance. J Appl Biomech Appl Physiol 90(5–6):549–553 9:95–95 Malliaras P, Kamal B, Nowell A, Farley T, Dhamu H, Simpson V, Erskine RM, Jones DA, Maganaris CN, Degens H (2009) In vivo spe- Morrissey D, Langberg H, Maffulli N, Reeves ND (2013) Patel- cific tension of the human quadriceps femoris muscle. Eur J Appl lar tendon adaptation in relation to load-intensity and contraction Physiol 106(6):827–838 type. J Biomech 46(11):1893–1899 Finni T, Ikegawa S, Lepola V, Komi P (2003) Comparison of force– Meylan CM, Cronin JB, Oliver JL, Hughes MG, McMaster D (2012) velocity relationships of vastus lateralis muscle in isokinetic The reliability of jump kinematics and kinetics in children of dif- and in stretch-shortening cycle exercises. Acta Physiol Scand ferent maturity status. J Strength Cond Res 26(4):1015–1026 177(4):483–491 Murtagh CF, Vanrenterghem J, O’Boyle A, Morgans R, Drust B, Foster BP, Morse CI, Onambele GL, Williams AG (2014) Human Erskine RM (2017) Unilateral jumps in different directions: COL5A1 rs12722 gene polymorphism and tendon proper- a novel assessment of soccer-associated power? J Sci Med ties in vivo in an asymptomatic population. Eur J Appl Physiol Sport 20(11):1018–1023 114(7):1393–1402 Murtagh CF, Nulty C, Vanrenterghem J, O’Boyle A, Morgans R, Drust Freriks B, Hermens H, Disselhorst-Klug C, Rau G (1999) The rec- B, Erskine RM (2018) The neuromuscular determinants of unilat- ommendations for sensors and sensor placement procedures for eral jump performance in soccer players are direction specific. Int surface electromyography. Seniam 8:13–54 J Sports Physiol Perform 13(5):604–611 Fukashiro S, Besier TF, Barrett R, Cochrane J, Nagano A, Lloyd DG Nagano A, Komura T, Fukashiro S (2004) Effects of the length ratio (2005) Direction control in standing horizontal and vertical jumps. between the contractile element and the series elastic element J Sport Health Sci 2005:272–279 on an explosive muscular performance. Electromyogr Kinesiol Grimshaw P, Fowler N, Lees A, Burden A (2004) BIOS instant notes 14(2):197–203 in sport and exercise biomechanics. Garland Science, New York Onambélé GN, Burgess K, Pearson SJ (2007) Gender-specific in vivo Guadagnoli E, Velicer WF (1988) Relation to sample size to the stabil- measurement of the structural and mechanical properties of the ity of component patterns. Psychol Bull 103(2):265 human patellar tendon. J Orthop Res 25(12):1635–1642 Hansen P, Bojsen-Moller J, Aagaard P, Kjær M, Magnusson SP (2006) Pearson SJ, Onambélé GL (2012) Computation methods affect the Mechanical properties of the human patellar tendon, in vivo. Clin reported values of in vivo human tendon stiffness. J Mech Behav Biomech 21(1):54–58 Mater 5(1):291–297 Helland C, Bojsen-Møller J, Raastad T, Seynnes O, Moltubakk M, Peltonen J, Cronin NJ, Stenroth L, Finni T, Avela J (2012) Achilles Jakobsen V, Visnes H, Bahr R (2013) Mechanical properties of tendon stiffness is unchanged one hour after a marathon. J Exp the patellar tendon in elite volleyball players with and without Biol 215(20):3665–3671 patellar tendinopathy. Br J Sports Med 47(13):862 Reed CC, Iozzo RV (2002) The role of decorin in collagen fibrillogen- Jaric S, Mirkov D, Markovic G (2005) Normalizing physical per- esis and skin homeostasis. Glycoconj J 19(4–5):249–255 formance tests for body size: a proposal for standardization. J Reeves ND, Narici MV (2003) Behavior of human muscle fascicles Strength Cond Res 19(2):467–474 during shortening and lengthening contractions in vivo. J Appl Kastelic J, Palley I, Baer E (1980) A structural mechanical model for Physiol 95(3):1090–1096 tendon crimping. J Biomech 13(10):887–893 Reeves ND, Maganaris CN, Narici MV (2003) Effect of strength train- Kawakami Y, Muraoka T, Ito S, Kanehisa H, Fukunaga T (2002) ing on human patella tendon mechanical properties of older indi- In  vivo muscle fibre behaviour during counter-movement viduals. J Physiol 548(3):971–981 1 3 European Journal of Applied Physiology Reeves ND, Maganaris CN, Ferretti G, Narici MV (2005) Influence of Stenroth L, Cronin NJ, Peltonen J, Korhonen MT, Sipilä S, Finni T 90-day simulated microgravity on human tendon mechanical prop- (2015) Triceps surae muscle–tendon properties in older endur- erties and the effect of resistive countermeasures. J Appl Physiol ance-and sprint-trained athletes. J Appl Physiol 120(1):63–69 98(6):2278–2286 Tillin NA, Jimenez-Reyes P, Pain MT, Folland JP (2010) Neuromus- Reilly T, Thomas V (1976) A motion analysis of work-rate in different cular performance of explosive power athletes versus untrained positional roles in professional football match-play. J Hum Mov individuals. Med Sci Sports Exerc 42(4):781–790 Stud 2(2):87–97 Wang H, Hua C, Cui H, Li Y, Qin H, Han D, Yue J, Liang C, Yang Rosager S, Aagaard P, Dyhre-Poulsen P, Neergaard K, Kjaer M, Mag- R (2013) Measurement of normal patellar ligament and anterior nusson S (2002) Load-displacement properties of the human tri- cruciate ligament by MRI and data analysis. Expert Ther Med ceps surae aponeurosis and tendon in runners and non-runners. 5(3):917–921 Scand J Med Sci Sports 12(2):90–98 Wiesinger H-P, Kösters A, Müller E, Seynnes OR (2015) Effects of Seynnes OR, Erskine RM, Maganaris CN, Longo S, Simoneau EM, increased loading on in  vivo tendon properties: a systematic Grosset J-F, Narici MV (2009) Training-induced changes in struc- review. Med Sci Sports Exerc 47(9):1885 tural and mechanical properties of the patellar tendon are related Wiesinger H-P, Rieder F, Kösters A, Müller E, Seynnes OR (2016) to muscle hypertrophy but not to strength gains. J Appl Physiol Are sport-specific profiles of tendon stiffness and cross-sectional 107(2):523–530 area determined by structural or functional integrity? PLoS One Seynnes O, Koesters A, Gimpl M, Reifberger A, Niederseer D, Nie- 11(6):e0158441 bauer J, Pirich C, Müller E, Narici M (2011) Effect of alpine Wiesinger H-P, Rieder F, Kösters A, Müller E, Seynnes OR (2017) skiing training on tendon mechanical properties in older men and Sport-specific capacity to use elastic energy in the patellar and women. Scand J Med Sci Sports 21(s1):39–46 achilles tendons of elite athletes. Front Physiol 8:132 Stafilidis S, Arampatzis A (2007) Muscle–tendon unit mechanical and Withers R, Maricic Z, Wasilewski S, Kelly L (1982) Match analysis of morphological properties and sprint performance. J Sports Sci Australian professional soccer players. Hum Mov Stud 8:159–176 25(9):1035–1046 Stenroth L, Peltonen J, Cronin NJ, Sipilä S, Finni T (2012) Age-related differences in Achilles tendon properties and triceps surae muscle architecture in vivo. J Appl Physiol 113(10):1537–1544 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Applied Physiology Springer Journals

Patellar tendon properties distinguish elite from non-elite soccer players and are related to peak horizontal but not vertical power

Free
13 pages
Loading next page...
 
/lp/springer_journal/patellar-tendon-properties-distinguish-elite-from-non-elite-soccer-0cm6q83K0q
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Biomedicine; Human Physiology; Occupational Medicine/Industrial Medicine; Sports Medicine
ISSN
1439-6319
eISSN
1439-6327
D.O.I.
10.1007/s00421-018-3905-0
Publisher site
See Article on Publisher Site

Abstract

Purpose To investigate potential differences in patellar tendon properties between elite and non-elite soccer players, and to establish whether tendon properties were related to power assessed during unilateral jumps performed in different directions. Methods Elite (n = 16; age 18.1 ± 1.0 years) and non-elite (n = 13; age 22.3 ± 2.7 years) soccer players performed vertical, horizontal-forward and medial unilateral countermovement jumps (CMJs) on a force plate. Patellar tendon (PT) cross- sectional area, elongation, strain, stiffness, and Young’s modulus (measured at the highest common force interval) were assessed with ultrasonography and isokinetic dynamometry. Results Elite demonstrated greater PT elongation (6.83 ± 1.87 vs. 4.92 ± 1.88 mm, P = 0.011) and strain (11.73 ± 3.25 vs. 8.38 ± 3.06%, P = 0.009) than non-elite soccer players. Projectile range and peak horizontal power during horizontal-forward CMJ correlated positively with tendon elongation (r = 0.657 and 0.693, P < 0.001) but inversely with Young’s modulus (r = − 0.376 and − 0.402; P = 0.044 and 0.031). Peak medial power during medial CMJ correlated positively with tendon elongation (r = 0.658, P < 0.001) but inversely with tendon stiffness (r = − 0.368, P = 0.050). Conclusions Not only does a more compliant patellar tendon appear to be an indicator of elite soccer playing status but it may also facilitate unilateral horizontal-forward and medial, but not vertical CMJ performance. These findings should be consid- ered when prescribing talent selection and development protocols related to direction-specific power in elite soccer players. Keywords Patellar tendon · Elongation · Strain · Stiffness · Young’s modulus · Unilateral · Countermovement jump AbbreviationsPT Patellar tendon BFlh Biceps femoris long head PCSA Ph ysiological cross-sectional area CMJ Countermovement jump Peak H-power P eak horizontal power normalised to 2/3 CSA Cross-sectional areabody mass sCSA Cr oss-sectional area normalised to body Peak M-power P eak medial power normalised to body 2/3 2/3 mass mass EMG Electromyography Peak V-power P eak vertical power normalised to body 2/3 mass RMVC R amped maximal voluntary contraction Communicated by Olivier Seynnes. VL Vastus lateralis * Conall F. Murtagh C.F.Murtagh@ljmu.ac.uk Introduction School of Sport and Exercise Sciences, Liverpool John Moores University, Liverpool L3 3AF, UK During the course of a match, an elite soccer player may Liverpool Football Club, Liverpool, UK perform up to ~ 119 maximal accelerations, ~ 35 sprints Department of Rehabilitation Sciences, KU Leuven, (Bradley et al. 2010), ~ 50 forceful changes of direction University of Leuven, 3000 Leuven, Belgium (Withers et  al. 1982), and ~ 16 vertical jumps (Reilly Football Association of Wales, Cardiff, Wales, UK and Thomas 1976). This activity profile implies that the Institute of Sport, Exercise and Health, University College most common type of powerful actions elicited involve London, London, UK Vol.:(0123456789) 1 3 European Journal of Applied Physiology the successive combination of eccentric and concentric performance is inconclusive. Moreover, the methods uti- muscle actions, also known as the stretch-shortening cycle lised in the previous research investigating the association (Cavanagh and Komi 1979). Indeed, elite soccer players between tendon properties and sport-specific performance have been previously shown to outperform non-elite soc- were limited, as they measured the properties of the VL cer players during unilateral jumping activities in different aponeurosis rather than the patellar tendon and, therefore, directions, which induce stretch-shortening cycle actions had to apply a series of corrections to estimate tendon in the muscle–tendon unit of the lower limb (Murtagh elongation (Kubo et al. 2011; Bojsen-Møller et al. 2005; et  al.  2017). The viscoelastic properties of the tendon Stafilidis and Arampatzis 2007). Assessing in vivo human affect the interaction between the contractile and elastic tendon properties directly via ultrasound and isokinetic elements of the muscle–tendon unit, and are thought to dynamometry (Seynnes et al. 2009; Hansen et al. 2006; influence performance during stretch-shortening cycle Wiesinger et al. 2016), on the other hand, would enable activities (Bojsen-Møller et al. 2005). However, the impor- a more precise investigation of a potential relationship tance of tendon properties for determining elite soccer between tendon properties and sport-specific measures of playing status, and their contribution to soccer-associated physical performance. power capabilities, has not yet been investigated. As the vast majority of explosive actions performed dur- A comparison of tendon properties in populations of ing elite soccer match-play involve a single legged push-off well trained and untrained, or different sporting perfor - (Bradley et al. 2010; Withers et al. 1982), and unilateral mance levels, may provide an insight into the importance CMJs in different directions differentiate elite from non- of these characteristics for high-level sports performance elite soccer players (Murtagh et al. 2017), such assessments (Tillin et al. 2010). Within this context, high- (Kubo et al. can be used to assess soccer-associated power (Murtagh 2011) and intermediate-level (Kubo et al. 2000b) sprinters et al. 2017). Indeed, we have previously shown that multi- have a more compliant vastus lateralis (VL) aponeurosis directional unilateral CMJs assess independent power quali- than untrained individuals. More specifically, significant ties (Murtagh et al. 2017) and are underpinned by discrete correlations have been reported between 100 m race per- neuromuscular factors (Murtagh et al. 2018). However, the formance and VL aponeurosis compliance (Kubo et al. importance of tendon properties in elite soccer and their 2000b) and maximal elongation (Stafilidis and Arampatzis contribution to unilateral CMJ performance in different 2007). This suggests that greater elongation of the VL ten- directions is unknown. Such information could inform the don–aponeurosis complex allows the utilisation of greater specific detail of soccer talent identification and develop - energy storage and increases the shortening velocity of the ment protocols relative to the different properties of the knee-extensor muscle–tendon unit during sprint running. PT. Therefore, the aims of this study were (1) to investigate Results from some vertical jump studies are in accordance whether PT properties differed between elite and non-elite with these findings, and report that VL aponeurosis stiff- soccer players; and (2) to establish if relationships existed ness correlated inversely with the calculated difference between the morphological, mechanical, and material prop- in jump height between bilateral vertical countermove- erties of the PT and unilateral jump performance in different ment jumps and squat jumps. These data suggest that the directions in soccer players. greater compliance of the VL tendon–aponeurosis com- plex facilitates the augmentation in jump performance with a countermovement (i.e., when comparing bilateral Methods vertical countermovement jumps vs. squat jumps) (Kubo et al. 1999). In contrast, Bojsen-Møller et al. (2005) dem- Subjects onstrated that VL aponeurosis stiffness correlated posi- tively with bilateral vertical countermovement jump height Twenty-nine male soccer players volunteered to take part in and knee-extensor isometric rate of torque development, this study, which was approved by Liverpool John Moores thus suggesting that a stiffer tendon–aponeurosis complex University Ethics Committee and complied with the Decla- contributes to enhanced muscle output during high force ration of Helsinki. Participants provided written informed isometric and dynamic bilateral vertical jumping tasks consent prior to being assigned to two groups according (Bojsen-Møller et  al. 2005). In agreement, Wiesinger to their level of competition. The elite soccer player group et al. (2016) directly assessed the properties of the patel- (n = 16; age 18.3 ± 1.3  years; height 1.81 ± 0.07  m; body lar tendon in vivo and showed significantly greater tendon mass 76.2 ± 9.7 kg) consisted of players from an English stiffness in elite ski jumping athletes compared to controls. Premier League football academy, who regularly partici- However, Wiesinger et  al. (2016) did not assess sport- pated at U18- and U21-level training and matches. The aver- specific performance directly and the previous research age weekly training/match content for elite soccer players investigating the contribution of tendon properties to jump was: 1 ± 1 soccer matches, 4 ± 1 soccer training sessions, 1 3 European Journal of Applied Physiology 3 ± 1  gym-based resistance training sessions (two upper Methodology body and one lower body), and 4 ± 1 non-resistance train- ing gym sessions (consisting of injury prevention, muscle Countermovement jumps activation and movement training). For more detail on the elite soccer player training frequency/type details, please see On arrival at the laboratory for the second session, all par- our recent study (Brownlee et al. 2018). The average weekly ticipants had their height and body mass measured. Partici- training content for the non-elite soccer players was: 1 ± 1 pants performed three trials of each CMJ type (with 60 s soccer matches, 2 ± 1 soccer training sessions, 1 ± 1 gym- recovery between trials within a single CMJ type, and 180 s based resistance training sessions, and no player reported between jump types), thus performing a total of 18 CMJs participating in any non-resistance training gym ses- (9 unilateral jumps on each leg). The methods for the per- sions. The inclusion criteria for the non-elite soccer player formance and data analysis of unilateral CMJs have been group (n = 13; age 22.4 ± 1.7  years; height 1.74 ± 0.06  m; explained in detail previously (Murtagh et al. 2017). As body mass 72.6 ± 6.6 kg) was to participate in at least 1 h recommended by Jaric et al. (2005), peak power outputs per week of competitive soccer (11-a-side or five-a-side), achieved during CMJs were isometrically scaled to body 0.67 and 1 h per week of soccer-specific or fitness-based training. mass (peak power divided by body mass ). The key per- Non-elite participants were excluded if they did not meet formance variables for the unilateral vertical CMJ were jump these inclusion criteria or had previously played soccer at height [calculated from the impulse–momentum relationship academy, semi-professional, or professional level. All par- derived take off velocity and equation of constant accelera- ticipants had been free of any injury to the lower body within tion methods (Dowling and Vamos 1993)] and peak vertical 0.67 the previous 3 months and had not previously sustained a power isometrically scaled to body mass (peak V-power). serious knee or ankle injury that may be aggravated dur- The key performance variables for horizontal-forward and ing testing procedures, or cause an adverse effect on perfor - medial CMJs were projectile range [calculated using equa- mance. Participants were fully familiarised with all testing tions of constant acceleration (Grimshaw et al. 2004)], peak procedures in a separate session and were asked to complete V-power, and peak horizontal power isometrically scaled to a physical activity and health questionnaire prior to the study body mass [(peak H-power) for horizontal-forward CMJs for screening purposes. This questionnaire allowed us to only] or peak medial power isometrically scaled to body ascertain if each potential participant satisfied the specific mass [(peak M-power) for medial CMJs only]. Projectile inclusion and exclusion criteria. range was used as the criterion performance measure for horizontal-forward and medial CMJs as, unlike when meas- Experimental design uring jump distance using a measuring tape, projectile range is not affected by airborne and landing technique and better All participants attended the laboratory on two occasions, represents the propulsive phase of the jump (Meylan et al. separated by between 3 and 7 days. The r fi st session enabled 2012). the participants to be familiarised with the assessment proto- col, which consisted of three unilateral CMJs in the vertical, Tendon morphology horizontal-forward, and medial directions on each leg, and two knee-extensor ramp maximum voluntary contractions Patellar tendon length, cross-sectional area (CSA), and elon- (RMVCs) on an isokinetic dynamometer. All CMJs were gation measurements were performed using ultrasonography visually demonstrated to the participants by the investiga- (MyLab30, Esaote, Genoa, Italy) with the knee joint set at tor. This session was also used to determine the superior 90° knee flexion (full extension = 0°) and the hip joint set at jumping leg (defined as the limb that produced the high- 85° (supine = 180°). The PT resting length, defined as the est ground reaction force during a unilateral vertical CMJ). distance between the patella apex and the point at which the During the second session, the participants performed all tendon inserts into the tibial tuberosity, was determined by assessments for CMJ and tendon properties on the dominant positioning the 40-mm-wide, 10–15 MHz linear transducer leg, and the results from these assessments were used for in the sagittal plane over the PT and marking the location subsequent analysis. To minimise the influence of previous of the patella apex (0% tendon length) and tibial tuberosity activity, the testing was performed at least 48 h following (100% tendon length) on the skin with a permanent marker any high-intensity exercise. Participants were also instructed pen. Three more locations were then marked on the skin over not to consume alcohol in the 48 h before testing sessions, the tendon (25, 50, and 75% tendon length) to enable PT and to consume no more than 3 mg/kg caffeine (250 mg as CSA to be measured from these axial images using Image an absolute caffeine guideline, e.g., no more than one cup of Analysis Software (ImageJ V.1.45s, National Institute of coffee) in the 5 h prior to the beginning of testing sessions. Health, MD, USA). Transverse scans were taken at each segment until an acceptable scan, in which the borders of the 1 3 European Journal of Applied Physiology whole PT cross section could be clearly identified. This was much from the required linear pattern upon visual inspec- generally achieved between one and three attempts. As the tion (Helland et al. 2013). Torque data during the RMVC stress imposed upon the PT during soccer activity is influ- were sampled using data acquisition software (AcqKnowl- enced by the forces exerted by the quadriceps femoris mus- edge, Biopac Systems Inc., Goleta, CA, USA). Ultrasound cle, in the absence of a published isometric power law for video sequences were recorded at 25 Hz during the RMVC this variable, PT CSA measurements were also isometrically and were synchronised with the RMVC torque data via the scaled to body mass (tendon cross-sectional area divided by administration of a square wave pulse, which was visible 0.67 body mass ; sCSA) (Seynnes et al. 2011). simultaneously on the AcqKnowledge software and the ultrasound monitor (ECG signal). It is acknowledged that Tendon elongation during RMVC this technique may lead to an underestimation of the PT elongation due to unmonitored tibial movements (Onambélé Details and reliability of the measurements have been et al. 2007). Unfortunately, the 4-cm-wide ultrasound trans- documented previously (Seynnes et al. 2009; Hansen et al. ducer did not enable simultaneous scanning of both proximal 2006). In short, the mechanical properties of the tendon were and distal insertions of the PT. However, because the major assessed by measuring the PT elongation during RMVCs. portion of the tendon was scanned in the same way in all par- Prior to performing two RMVCs, a 2-mm-wide strip of sur- ticipants, this technical compromise does not invalidate the gical tape (3M, Neuss, Germany) acted as an echo-absorbent comparative outcome of the data. Furthermore, this method marker and was placed on the skin transversely over the of assessing PT elongation has been shown to have a good tendon as a reference point at ~ 30% tendon length from the test–retest reproducibility by us (Table 1) and others (Reeves proximal end. Immediately before the RMVCs, participants et al. 2003). performed a series of five submaximal isometric contrac - tions to ensure preconditioning of the tendon. The linear Antagonist muscle co-activation transducer was then placed in the sagittal plane over the PT (~ 1 cm of the probe above the patellar apex and ~ 3 cm Antagonist (hamstring) muscle co-activation during RMVC below, incorporating the echo-absorbent marker). Similar to was assessed via electromyography (EMG) activity. The the previous research that prescribed a specific time period biceps femoris long head (BFlh) muscle [representative for ramped isometric contractions when measuring tendon of the knee flexor muscle group (Kellis and Baltzopoulos properties (Hansen et al. 2006; Seynnes et al. 2009; Mal- 1997)] was identified via palpation during a submaximal liaras et al. 2013), the RMVCs lasted 6 s in total with 2 min knee flexion with the participant in the prone position. rest between each contraction. Loading rates (Nm/s) were, After preparing the skin surface (shaving, lightly abrad- therefore, dependent on the participant’s maximal volun- ing, and cleansing with 70% ethanol) at 2/3 muscle length tary force capacity. Visual feedback of force production was from the proximal end [SENIAM guidelines (Freriks et al. displayed on a screen in front of the participants to ensure 1999)], two bipolar Ag–AgCl surface electrodes (Neu- that all RMVCs were performed at a constant loading rate. roline, Medoicotest, Rugmarken, Denmark) were placed Trials were discarded when the torque trace deviated too 20  mm apart along the sagittal axis of the muscle belly Table 1 Test–retest Tendon property Typical error (95% CI) CV (%) ICC (95% CI) reproducibility of patellar tendon measurements in eight CSA (25% TL) 1.460 (0.966–2.972) mm 1.234 0.997 (0.987–0.999) healthy, recreationally active CSA (50% TL) 1.685 (1.114–3.429) mm 1.694 0.978 (0.898–0.995) young men CSA (75% TL) 2.591 (1.713–5.273) mm 1.996 0.988 (0.941–0.998) Mean CSA 1.190 (0.787–2.421) mm 1.004 0.996 (0.980–0.999) Stress 0.4065 (0.269–0.827) MPa 0.968 0.996 (0.982–0.999) Elongation 0.322 (0.213–0.657) mm 5.653 0.986 (0.935–0.997) Strain 0.704 (0.465–1.432) % 5.506 0.986 (0.936–0.997) Stiffness 63.6 (39.7–156.1) N/mm 4.477 0.988 (0.912–0.998) Young’s modulus 0.032 (0.021–0.071) GPa 5.930 0.979 (0.889–0.996) Measurements were taken on two separate occasions (at the same time of day) within a 12-day period. CSA measurements were taken at rest with the knee flexed at 90° (0° = full extension). Tendon stress, elonga- tion, strain, stiffness, and Young’s modulus were all calculated using the highest common force (interval) CV coefficient of variation, ICC intra-class correlation coefficient, CSA cross-sectional area, TL tendon length, MPa megapascals, GPa gigapascals P < 0.001 for all ICCs 1 3 European Journal of Applied Physiology and a reference electrode placed on the lateral tibial con- However, we do acknowledge that this technique is limited, dyle. The root mean square (RMS) of the BFlh EMG signal as we are assuming that the force elongation relationship was recorded at each 10% knee extension R MVC and for between 2599–3248  N is linear (Pearson and Onambélé 500 ms around the peak torque during a knee flexion MVC 2012). PT elongation was calculated as the change in tendon at 90° knee flexion. Assuming a linear relationship between length at (3248 N) from resting length. Tendon strain was EMG and torque output (Kellis and Baltzopoulos 1997), calculated as the change in tendon length at the highest com- the torque generated by the antagonists during each 10% mon RMVC force level relative to the original tendon length knee extension RMVC was estimated by dividing the BFlh (∆L/L ), and expressed as a percentage. Young’s modulus EMG during knee-extensor RMVC by the BFlh maximal (E) was calculated by multiplying stiffness (k ) with the ratio EMG (BFlh EMG during knee flexion MVC) and multiply - of the resting tendon length (l ) to mean tendon CSA, i.e., ing this ratio by the knee flexion MVC torque at each 10% E = k × (l /s). RMVC. This torque was added to the knee-extensor torque value at the relevant 10% RMVC to provide the gross knee- Test–retest reproducibility extensor RMVC torque. Dividing the RMVC torque by the PT moment arm at 90° knee flexion [assumed to be 0.048 m To establish the test–retest reproducibility of our methods based on MRI measurements in young healthy men (Erskine for measuring the morphological, mechanical, and material et al. 2009)] provided the PT force at each 10% RMVC. Due properties of the PT, we recruited an additional eight rec- to technical issues with some EMG recordings, EMG data reationally active, young, healthy men (age 28.0 ± 3.0 years; were only available in a subsample of elite (n = 12) and non- body mass 77.1 ± 7.4 kg; height 1.79 ± 0.04 m). All par- elite (n = 8) participants. As a limited sample size will nega- ticipants attended the laboratory on two separate occasions, tively impact the statistical power of correlation analyses at the same time of day, in the same standardised condi- (Guadagnoli and Velicer 1988), and investigating the rela- tions as specified for the elite and non-elite soccer players, tionship between tendon properties and power during CMJs with a period of 7–12 days in between each testing session. was one of the main aims of this study, we have reported Measurements were performed by the same researcher, as tendon properties without accounting for antagonist muscle described in the “Tendon morphology”, “Tendon elonga- co-activation [which is not uncommon practice (Stenroth tion during RMVC”, and “Analysis of tendon data” sections. et al. 2012, 2015; Peltonen et al. 2012)] in an effort to max- Inter-day reliability for each measurement was expressed as imise statistical power. However, to support these findings, typical error (TE), coefficient of variation (CV), and intra- we have also reported the results from correlation analyses class correlation coefficient (ICC, model: two-way mixed; in the subsample, where antagonist muscle co-activation type: absolute agreement) with 95% confidence intervals was accounted for. (CIs). These data are presented in Table  1, and for all variables, the TE and CV were low, and the ICC was high Analysis of tendon data (≥ 0.979) with narrow CIs, thus indicating that our methods were highly reproducible. PT elongation was recorded as the distance the patella apex moved from the external marker every 10% RMVC. Statistical analyses The video frame that corresponded to each 10% RMVC was exported as a portable network graphics (.png) file and The mean and standard deviation (s) were calculated for all the distance from the patella apex to the external marker variables in both the whole cohort, and a subsample where was measured using image analysis software (ImageJ v. antagonist co-activation was accounted for. All data were 1.47, National Institute of Health, MD, USA). Individual tested for normality using the Shapiro–Wilks normality test. force–elongation curves were fitted with a second-order For tendon CSA, tendon sCSA, and tendon elongation at polynomial (R > 0.95 in all cases). Tendon stiffness each 10% RMVC interval, a two-way mixed ANOVA was (∆Force/∆PT length) values were obtained over the high- used to determine a main effect of athlete status (between est 20% common force interval in the weakest participant factor: elite vs. non-elite soccer players), tendon location with the lowest maximal RMVC peak force (2599–3248 N). [within factor: CSA at 0, 25, 50, 75, and 100% tendon length This approach, as performed by others (e.g., Seynnes et al. (this includes mean tendon CSA comparison)] or tendon 2009), was necessary to avoid having to extrapolate some elongation (within factor: tendon elongation at 10, 20, 30, data points beyond the visible part of the force–elonga- 40, 50, 60, 70, 80, 90, and 100% RMVC), and an athlete tion curve relationship (which would have been required status × tendon location, or athlete status × tendon elonga- for the weaker participants). It also allowed us to compare tion at a specific % of RMVC, interaction. If a significant the mechanical and material properties of elite and non- interaction occurred, simple main effects and pairwise elite player tendons under the same forces and conditions. comparisons with Bonferroni adjustment were performed 1 3 European Journal of Applied Physiology to reveal differences. If a significant main effect existed for tendon CSA location or tendon elongation at a specific % of RMVC, Bonferroni post hoc tests were used to establish the difference in CSA or tendon elongation between tendon locations or % RMVC force levels, respectively. For tendon force, length, stiffness, elongation, strain, and Young’s mod- ulus, independent t tests were used to determine differences between elite and non-elite soccer players, in both the whole cohort and subsample (where antagonist co-activation was measured). Pearson’s correlations were used to determine relationships between jump performance variables (height or projectile range, peak V-power, peak H-power, or peak M-power) and tendon mean CSA and CSA at 0, 25, 50, 75, Fig. 1 Patellar tendon cross-sectional area (CSA) along its length at 25% intervals in elite (n = 16, circles, mean + SD) and non-elite and 100% tendon length, strain, stiffness, and Young’s mod- (n = 13, open squares, mean − SD) players. *Main effect, elite players ulus. Statistical analysis was completed using SPSS version significantly greater CSA than non-elite players (P < 0.05) 21 (SPSS Inc., Chicago, IL, USA) and the significance level was set at P ≤ 0.05. Table 2 Patellar tendon morphological, mechanical, and material properties calculated in the whole cohort [elite (n = 16) and non-elite (n = 13)]; mean ± SD Results Tendon property Elite Non-elite Anthropometry Mean CSA (mm ) 115 ± 6* 112 ± 3 2 0.67 Mean sCSA (mm /kg ) 6.35 ± 0.44 6.41 ± 0.02 Elite soccer players were significantly taller (P = 0.009) Resting tendon length (mm) 58.5 ± 7.1 58.2 ± 4.4 and had significantly longer femur length (46.8 ± 2.1 vs. Stiffness (N/mm) 1269 ± 607 1707 ± 792 44.0 ± 1.7  cm; P = 0.001) than non-elite. However, there Elongation (mm) 6.83 ± 1.87* 4.92 ± 1.88 was no difference in body mass between elite and non-elite Strain (%) 11.73 ± 3.25** 8.38 ± 3.06 soccer players (P = 0.267). Stress (MPa) 28.29 ± 1.34* 29.15 ± 0.72 Young’s modulus (GPa) 0.64 ± 0.31 0.88 ± 0.39 Differences in patellar tendon properties MVC tendon force (N) 5728 ± 1522 4800 ± 989 between elite and non‑elite players CSA cross-sectional area, sCSA cross-sectional area isometrically 0.67 scaled to body mass , MVC maximal voluntary contraction Tendon morphology *A significant difference when comparing elite and non-elite (P < 0.05) For tendon CSA, there was a main effect of athlete sta- **A significant difference between elite and non-elite (P < 0.01) tus, with elite soccer players demonstrating ~ 2.6% greater mean CSA (F = 4.439, P = 0.045; Fig. 1; Table 2). There 1,27 was also a main effect of tendon location ( F = 105.36, 0 vs. 75%, P < 0.001; 0 vs. 100%, P < 0.001; 25 vs. 75%, 4,108 P < 0.001) with pairwise post hoc analyses, revealing that P < 0.001; 25 vs. 100%, P < 0.001; 50 vs. 75%, P < 0.001; tendon CSA varied along the length of the tendon (0 vs. 50 vs. 100%, P < 0.001; 75 vs. 100%, P = 0.031); except at 25%, P = 0.001; 0 vs. 50%, P < 0.001; 0 vs. 75%, P < 0.001; 25 vs. 50% tendon length where there was a non-significant 0 vs. 100%, P < 0.001; 25 vs. 75%, P < 0.001; 25 vs. 100%, difference (P = 0.185). However, there was no significant 0.67 P < 0.001; 50 vs. 75%, P < 0.001; 50 vs. 100%, P < 0.001; main effect of athlete status (elite 0.063 ± 0.004 mm/kg ; 0.67 75 vs. 100%, P = 0.007; Fig. 1); except at 25 vs. 50% ten- non-elite 0.063 ± 0.004 mm/kg , F = 0.006, P = 0.941) 1,27 don length where there was a non-significant difference and no significant interaction between athlete status and ten- (P = 0.063; Fig.  1). There was no significant interaction don CSA location (F = 0.720, P = 0.580). Furthermore, 4,108 between tendon CSA and athlete status (F = 0.720, tendon resting length relative to femur length did not differ 4,96 P = 0.565). Tendon resting length did not differ between between groups (t = − 1.520; P = 0.140). groups (t = 0.119; P = 0.906; Table 2). For tendon sCSA, there was a main effect of tendon loca - Tendon mechanical and material properties tion (F = 106.07, P < 0.001) with pairwise post hoc 4,108 analyses, revealing that tendon CSA varied along the length We present differences in the tendon mechanical and mate- of the tendon (0 vs. 25%, P = 0.024; 0 vs. 50%, P < 0.001; rial properties between elite and non-elite soccer players 1 3 European Journal of Applied Physiology calculated in both the whole cohort (elite: n = 16; non-elite: Relationships between patellar tendon properties n = 13) and subsample (elite: n = 12; non-elite, n = 8) where and direction‑specific jump performance antagonist co-activation was accounted for. Elite soccer players demonstrated significantly greater tendon elonga - The positive and inverse relationships between jump per- tion and strain than non-elite players (whole cohort: elonga- formance variables and tendon properties from the whole tion, P = 0.011; strain, P = 0.009; Table 2; subsample: elon- cohort are displayed in Table 3. Additional gu fi res have been gation, 6.72 ± 1.94 vs. 4.76 ± 1.97  mm, P = 0.041; strain, included to illustrate the spread of data for specific relation - 11.38 ± 3.27 vs. 8.03 ± 3.25%, P = 0.037). Moreover, when ships. In the subsample, correlations between jump perfor- we compared tendon elongation at each 10% RMVC inter- mance variables and tendon properties are also illustrated val in the whole cohort, there was a significant main effect in the text. of athlete status (F = 93.00, P < 0.001, Fig. 2) and tendon In the whole cohort and subsample, unilateral horizontal- 9,243 elongation (F = 11.13, P < 0.001, Fig. 2), but no significant forward jump performance variables correlated positively 9,27 interaction effect between tendon elongation and athlete sta- with tendon elongation (whole cohort: Fig.  3a; Table  3; tus (F = 1.63, P = 0.108). This finding suggests that elite subsample: projectile range, r = 0.733, P < 0.001, peak 9,243 soccer player tendon elongation was greater than non-elite H-power: r = 0.734, P < 0.001) and strain (whole cohort: soccer players at all 10% levels of RMVC. However, there Table 3; subsample: projectile range: r = 0.765, P < 0.001, were no differences between elite and non-elite soccer players peak H-power: r = 0.677, P = 0.001), but inversely with regarding tendon stiffness (whole cohort: P = 0.104; Table 2; tendon stiffness (whole cohort: Fig.  3b; Table 3; subsam- subsample: P = 0.496). There was a non-significant tendency ple: projectile range: r = − 0.576, P = 0.008, peak H-power: for Young’s modulus (P = 0.075; Table 2) to be lower and r = − 0.579, P = 0.007) and Young’s modulus (whole cohort: maximum tendon force (P = 0.069; Table 2) to be higher in Fig. 3c; Table 3; subsample: projectile range: r = − 0.559, elite compared to non-elite soccer players in the whole cohort. P = 0.010, peak H-power: r = − 0.490, P = 0.028). Simi- In the subsample, elite soccer players produced significantly larly, unilateral medial CMJ performance variables cor- greater maximum tendon force in comparison to non-elites related positively with tendon elongation (whole cohort: (6374 ± 1658 vs. 4817 ± 607 N, P = 0.021). Fig.  4a; Table  3; subsample: projectile range: r = 0.513, For a comparison of unilateral jump performance between P = 0.021, peak M-power: r = 0.633, P = 0.003) and strain elite and non-elite soccer players, please refer to our data (whole cohort: Fig. 4b; Table 3; subsample: peak M-power: presented in a previously published study specifically aimed r = 0.570, P = 0.009), but inversely with tendon stiffness to determine whether jump performance could distinguish (whole cohort: Fig. 4c; Table 3; subsample: tendency for between soccer playing levels (Murtagh et al. 2017). a significant correlation with peak M-power: r = − 0.432, P = 0.057). Mean tendon CSA, CSA at 75% and CSA at 100% correlated positively with unilateral horizontal-for- ward CMJ peak V-power and unilateral medial CMJ peak V-power (Table 3) in the whole cohort. Furthermore, tendon CSA at 50 and 100% correlated positive with unilateral ver- tical jump peak V-power (Table 3). There were no significant correlations between any jump variable and tendon CSA at 0 or 25% (r ≤ 0.307; P ≥ 0.116) (Table 3). Discussion The aims of our study were to compare PT properties between elite and non-elite soccer players, and to establish if these properties were related to unilateral jump perfor- mance in different directions. Our results showed that ten- don CSA, elongation and strain were greater in elite soccer 0 1 2 3 4 5 6 7 8 9 10 11 players. These specific tendon properties may, therefore, Elongation (mm) be considered determinants of elite soccer playing status. Tendon elongation and strain also correlated positively with Fig. 2 In vivo patellar tendon force–elongation relation in elite peak horizontal power and medial power during unilateral (n = 16; circles) and non-elite (n = 13; open squares) soccer play- horizontal and unilateral medial CMJ, respectively, while ers. Significant main effect for athlete status (P < 0.05). Data are mean + SD tendon stiffness and Young’s modulus correlated inversely 1 3 Tendon force (N) European Journal of Applied Physiology Table 3 Correlations between unilateral countermovement jump (CMJ) performance measures and patellar tendon properties in elite (n = 16) and non-elite (n = 13) soccer players Jump type Unilateral Vertical CMJUnilateral Horizontal CMJUnilateral Medial CMJ Tendon property Jump height Peak V-power Projectile range Peak H-power Peak V-power Projectile range Peak M-power Peak V-power (cm) (cm) (W/kg) (W/kg) (W/kg) (W/kg) (W/kg) (cm) Stiffness, N/mm r = -0.019 r = 0.168 r = -0.333 r = -0.409 r = 0.013 r = -0.150 r = -0.368 r = 0.207 P = 0.924 P = 0.393 P = 0.077 P = 0.028 P = 0.948 P = 0.436 P = 0.050 P = 0.282 Elongation, mm r = 0.114 r = -0.021 r = 0.657 r = 0.693 r = 0.067 r = 0.496 r = 0.658 r = 0.078 P = 0.473 P = 0.917 P < 0.001 P < 0.001 P = 0.731 P = 0.006 P < 0.001 P = 0.687 Strain, % r = 0.248 r = 0.066 r = 0.698 r = 0.674 r = 0.180 r = 0.409 r = 0.616 r = 0.097 P = 0.204 P = 0.740 P < 0.001 P < 0.001 P = 0.350 P = 0.027 P < 0.001 P = 0.615 Young’s modulus, GPa r = -.081 r = 0.109 r = -0.376 r = -0.402 r = -0.081 r = 0.089 r = -0.344 r = 0.195 P = 0.681 P = 0.583 P = 0.044 P = 0.031 P = 0.675 P = 0.648 P = 0.068 P = 0.311 Mean CSA, mm r = 0.165 r = 0.339 r = 0.109 r = -0.038 r = 0.381 r = -0.009 r = -0.020 r = 0.401 P = 0.402 P = 0.076 P = 0.078 P = 0.845 P = 0.040 P = 0.961 P = 918 P = 0.031 CSA at 50% tendon length r = 0.233 r = 0.376 r = 0.066 r = 0.006 r = 0.213 r = 0.120 r = 0.067 r = 0.308 mm P = 0.233 P = 0.049 P = 0.733 P = 0.975 P = 0.267 P = 0.534 P = 0.732 P = 0.104 CSA at 75% tendon length r = 0.024 r = 0.082 r = 0.251 r = 0.026 r = 0.396 r = -0.048 r = -0.036 r = 0.416 mm P = 0.904 P = 0.677 P = 0.190 P = 0.892 P = 0.033 P = 0.804 P = 854 P = 0.025 CSA at 100% tendon length r = 0.189 r = 0.414 r = 0.150 r = -0.013 r = 0.496 r = -0.040 r = -0.172 r = 0.654 mm P = 0.334 P = 0.029 P = 0.437 P = 0.948 P = 0.006 P = 0.837 P = 0.372 P < 0.001 Significant correlations are highlighted in grey. Significant inverse correlations are highlighted in bold 0.67 CSA cross-sectional area, peak V-power peak vertical power isometrically scaled to body mass , peak H-power peak horizontal power iso- 0.67 0.67 metrically scaled to body mass , peak M-power peak medial power isometrically scaled to body mass with horizontal power during unilateral horizontal CMJ. and tendon size (see Wiesinger et al. 2015 for review), thus Only PT CSA at 50% and 100% tendon length was related potentially suggesting that long-term exposure to elite soccer to unilateral vertical CMJ performance. Our study suggests training results in greater PT tendon CSA. However, when 0.67 that a more compliant PT facilitates greater unilateral CMJ tendon CSA was normalised to body mass , there was no performance in the horizontal-forward and medial direc- significant difference in tendon CSA between cohorts at any tions, but PT stiffness does not contribute to vertical CMJ tendon location (no significant main effect of athlete status performance. and no significant interaction between athlete status and ten- Comparing the physiological capabilities in soccer play- don CSA). Although we found no significant difference in 0.67 ers competing at different performance levels may provide body mass between the elite and non-elite players, this an insight into the physiological factors that determine approach was necessary due to the close relationship found elite soccer playing status. Tendon CSA was significantly previously between body mass and tendon CSA (Rosager greater in elite compared to non-elite soccer players at all et al. 2002). Elite players were significantly taller than non- tendon locations (main effect of athlete status and no sig- elites and the difference in tendon CSA between elite and nificant interaction between athlete status and tendon CSA). non-elite players may, therefore, be related to individual Although the contents of elite and non-elite player habitual differences in body size rather than adaptations to chronic weekly training schedules seem similar (please refer to the habitual loading patterns. Interestingly, in agreement with “Methods” section), elite soccer players perform a higher other literature (Wang et al. 2013), although elite players volume of soccer training sessions and, unlike non-elite displayed greater stature and femur length compared to non- players, also partake in non-resistance-based gym training. elites, there was no difference in resting tendon length, thus Elite players are, therefore, exposed to a greater number of potentially suggesting that the length of the PT in soccer cyclic lower limb loading cycles and high impact explosive players is not related to their body dimensions. Neverthe- actions per week, both of which have been shown to poten- less, without a control group (who did not partake in soccer- tially stimulate increments in tendon CSA (Magnusson and specific training) to compare with, or longitudinal training Kjaer 2003; Couppe et al. 2008; Wiesinger et al. 2016). Sub- study, it still remains unknown if long-term soccer training sequently, our results may agree with the previous literature stimulates PT CSA adaptation. documenting the relationship between habitual daily loading 1 3 European Journal of Applied Physiology Fig. 4 Relationship between unilateral medial (UM) counter- movement jump (CMJ) peak M-power and tendon elongation (a, Fig. 3 Relationship between unilateral horizontal (UH) countermove- r = 0.658, P < 0.001), strain (b, r = 0.616, P < 0.001), and stiffness (c, ment jump (CMJ) peak H-power and tendon elongation (a, r = 0.693, r = − 0.368, P = 0.050) in elite (n = 16, circles) and non-elite (n = 13, P < 0.001), stiffness (b, r = − 0.409, P = 0.028), and Young’s Modu- open squares) players. Peak M-power peak medial power isometri- lus (c, r = − 0.402, P = 0.031) in elite (n = 16, circles) and non-elite 0.67 cally scaled to body mass (n = 13, open squares) players. Peak H-power peak horizontal power 0.67 isometrically scaled to body mass Despite greater tendon CSA, tendon strain was higher in of the common highest 20% force interval of the weakest elite compared to non-elite soccer players. This illustrates participant: 2599–3248 N) were applied to the PT, there that elite soccer players have PT structures that elongate was no significant difference in the change in elongation more, relative to their original length, compared to non- between elite and non-elite players. It, therefore, appears elites at a given tendon force value (highest common ten- that the greater magnitude of difference in PT elongation don force = 3248 N). Although elite soccer players displayed between elite and non-elite players occurs at lower force greater PT elongation, we found no significant difference levels (below the common force interval of the weakest par- in tendon stiffness between elite and non-elite soccer play - ticipant). Young’s modulus (i.e., the relation between stress ers. This suggests that when specific forces (representative and strain), which represents the material properties of the 1 3 European Journal of Applied Physiology tendon independent of its dimensions, provides a more accu- unilateral horizontal-forward CMJ peak H-power and uni- rate representation of the in vivo function of the PT when lateral medial CMJ peak M-power, while tendon stiffness comparing elite and non-elite soccer players (Foster et al. was inversely related to unilateral horizontal-forward CMJ 2014). Our findings showed that there was a tendency for peak H-power and unilateral medial CMJ peak M-power. elite soccer players to display lower Young’s modulus than These findings suggest that more compliant PTs facilitate non-elite players. The greater elongation and strain dem- greater unilateral horizontal and unilateral medial CMJ onstrated by elite in comparison to non-elite soccer players performance in soccer players. Moreover, Young’s modu- may, therefore, be due to differences in the microstructure lus was inversely related with unilateral horizontal CMJ of the tendon, including less collagen cross linking and projectile range and peak H-power, thus suggesting that less fibril packing (Reed and Iozzo 2002). Moreover, as the a tendon with more elastic properties enhances unilateral major difference in the change in PT elongation between horizontal CMJ performance. elite and non-elite players occurs at lower force levels, it may As acceleration and sprinting activities require the pro- be that elite soccer players display PTs with more collagen duction of high levels of horizontal-forward power (Buch- crimping (Kastelic et al. 1980). However, given the non- heit et al. 2014), these findings are in accordance with the significant difference in Young’s modulus between groups, previous studies reporting that 100 m sprint performance is this theory remains speculative. positively related to VL aponeurosis compliance (Kubo et al. Considering the sprint demands of elite soccer match- 2000b) and maximal elongation (Stafilidis and Arampatzis play (Bradley et  al. 2010), the greater PT elongation in 2007). Real-time ultrasonography observations of tendon elite compared to non-elite soccer players in our study is behaviour in  vivo has shown that, during high-intensity consistent with previous studies, reporting that the VL ten- jumping movements whereby the range of joint motion is don–aponeurosis complex of elite sprinters elongates more small, the VL muscle fascicles lengthen only marginally, if than both non-elite sprinters (Stafilidis and Arampatzis at all during the eccentric phase (Finni et al. 2003), and are 2007) and untrained participants (Kubo et al. 2011). While thought to function quasi-isometrically. Horizontal-forward it has been well documented that resistance and plyometric CMJs have been shown to require ~ 10° less knee flexion training interventions increase PT stiffness (see Wiesinger than vertical CMJs (Fukashiro et al. 2005). Hence, unilateral et  al. 2015 for review), no specific training intervention horizontal-forward and possibly unilateral medial CMJs may has been shown to increase the elongation properties of require a quasi-isometric contraction of the knee-extensor the PT structures (Seynnes et al. 2009; Reeves et al. 2003), muscle group, inducing greater tendon lengthening (Reeves and only bed rest (Reeves et al. 2005), detraining (Kubo and Narici 2003), thus allowing the tendon to store more et al. 2010), and ageing (Reeves et al. 2003) are known to potential energy and recoil at greater speeds. This would induce these changes. Therefore, as the CC genotype of the enable the tendon to act as a power amplie fi r during horizon - COL5A1 (rs12722) genetic variant (which encodes the pro tal-forward and medial CMJs (Nagano et al. 2004). Moreo- alpha chain of the type V collagen) has previously associated ver, it has also been reported that during the initial concen- with greater PT elongation and lower stiffness (Kubo et al. tric phase of stretch-shortening cycle exercises, the rapid 2013), it is possible that elite soccer players have a genetic shortening of the tendon contributes to lowering the short- predisposition to tendon structures with greater elongation ening velocity of the muscle fibres to near isometric condi- properties compared to non-elite soccer players. However, tions (Kawakami et al. 2002). Therefore, a more compliant the research associating specific genetic variants and PT PT will have a capacity for greater elongation, allowing the properties remains inconclusive (Foster et al. 2014) and it knee-extensor muscle fibres more time to develop greater cannot be discounted that certain types or combinations of forces during the concentric propulsive phase of sprinting long-term soccer development training stimuli, which have (Kubo et al. 2011), unilateral horizontal-forward CMJ and not been investigated to date, could induce such tendon unilateral medial CMJ activities. Hence, our data support adaptations. Nevertheless, our findings highlight the poten- a notion that a more compliant PT could enhance soccer tial importance of assessing PT elongation in elite soccer player performance levels during activities that depend heav- talent selection and development protocols. ily upon horizontal push-off capacity. Unilateral CMJ tasks oriented in different directions While tendon CSA was not related to unilateral hori- have been suggested to represent a measurement of soccer- zontal-forward or medial CMJ peak H-power or M-power, associated power (Murtagh et al. 2017), and the relation- respectively, distal tendon CSA was positively associated ship between tendon properties and unilateral CMJ perfor- with the ability to produce peak vertical power during uni- mance in different directions are, therefore, of interest for lateral vertical, horizontal and medial CMJs (Table 3). How- the assessment and development of muscular power in elite ever, there was no relationship between any mechanical or soccer players. Our study showed that in soccer players, material PT properties with peak V-power during any CMJ. PT elongation and strain were both positively related to We have previously shown that unilateral vertical and medial 1 3 European Journal of Applied Physiology CMJ peak V-power is positively related to quadriceps femo- that long-term elite soccer training stimulates such adapta- ris muscle size (Murtagh et al. 2018). Considering the rela- tions and may support the hypothesis of functionally driven tionship between body size and tendon CSA (Rosager et al. tendon adaptation proposed by Wiesinger et  al. (2016). 2002), players with larger quadriceps may also have greater However, future research is needed to confirm this theory PT CSA. Hence, the ability to produce peak V-power may as when we accounted for body size, no significant differ - be more closely related to the neuromuscular properties of ence between elite and non-elite players was found. We also the knee extensors rather than the properties of the PT and show that a more compliant PT appears to facilitate horizon- this may be the reason why we found a positive relationship tal-forward and medial jump performance, and therefore, between tendon CSA and peak V-power. The lack of direct practitioners aiming to develop unilateral explosive perfor- relationships between PT mechanical and material proper- mance in these directions should, perhaps, reconsider the ties and unilateral vertical CMJ performance in the soccer prescription of training interventions that may induce a stiff- players in our study is in accordance with some (Kubo et al. ening of the PT. We recommend that future research should 2000a), but not all the previous studies (Bojsen-Møller et al. also investigate training interventions that may increase PT 2005). Bojsen-Møller et al. (2005) found that a stiffer VL elasticity. tendon–aponeurosis complex contributed to enhanced peak force and power outputs during high-intensity isometric and dynamic bilateral vertical jumping tasks, respectively. Conclusion Discrepancies between the findings of Bojsen-Møller et al. (2005) and our results are likely due to methodological dis- We have shown that patellar tendon CSA, elongation, and parities. Whilst we analysed ultrasound images of the PT, strain are greater in elite compared to non-elite soccer play- Bojsen-Møller et al. (2005) measured the VL aponeurosis ers, and are, therefore, important indicators of U18 and U21 and, therefore, approximated tendon elongation. Conse- elite soccer playing status. More compliant patellar tendon quently, it is possible that their stiffness measurements do properties were positively related to performance in jumps not account for total displacement of the tendon and may with a horizontal push-off, yet they were not related to ver - underestimate tendon compliance. tical jump performance. We show that a more compliant We do acknowledge some limitations to our study that patellar tendon will likely benefit elite soccer players across could inform future research. Our investigation only includes all soccer-specific activities involving a powerful single-leg elite players from one soccer club and should be replicated push off in the horizontal direction. in a wider range of soccer institutions. Moreover, we have Acknowledgements The authors wish to thank Raja Azidin and Chris not directly investigated the capacity of the tendon to store Nulty for their help and expertise during the lab testing procedures, and release energy in vivo and future research should also Remy Tang and Neil Critchley for their co-operation with the recruit- measure tendon hysteresis [as performed by Wiesinger ment of elite players, and the participants from Liverpool Football Club et al. (2017)]. The loading rates during the ramped isomet- Academy and Liverpool John Moores University. ric voluntary contractions were dependent on the partici- pants maximal voluntary force level. The previous literature Compliance with ethical standards shows that there are differences in the stiffness and Young’s Conflict of interest The authors declare that they have no conflict of modulus calculations when ramped isometric contractions interest. are performed at 50 N m/s compared to 80 and 110 N m/s (Kösters et al. 2014). Subsequently, we cannot discount that Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the insti- the different loading rates may have affected calculations tutional and/or national research committee and with the 1964 Helsinki of stiffness and Young’s modulus. However, as the loading declaration and its later amendments or comparable ethical standards. rates between participants in our study (29.7–71.3 N m/s) are within a narrow range, we believe it is unlikely that this Informed consent Written informed consent was obtained from all individual participants included in the study. method of assessment compromised the validity of our data. Moreover, as the previous studies adopted similar method- ologies (Hansen et al. 2006; Seynnes et al. 2009; Malliaras Open Access This article is distributed under the terms of the Crea- et al. 2013), we believe that our data are comparable to cur- tive Commons Attribution 4.0 International License (http://creat iveco mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- rent research studies. tion, and reproduction in any medium, provided you give appropriate Our study suggests that novel soccer talent selection pro- credit to the original author(s) and the source, provide a link to the tocols should include a measurement of PT properties, and Creative Commons license, and indicate if changes were made. elite soccer clubs should aim to recruit, and develop, play- ers with more elastic/compliant PTs. As elite players dis- played greater absolute tendon CSA, our research suggests 1 3 European Journal of Applied Physiology exercise in humans reveals a significant role for tendon elastic- References ity. J Physiol 540(2):635–646 Kellis E, Baltzopoulos V (1997) The effects of antagonist moment Bojsen-Møller J, Magnusson SP, Rasmussen LR, Kjaer M, Aagaard on the resultant knee joint moment during isokinetic test- P (2005) Muscle performance during maximal isometric and ing of the knee extensors. Eur J Appl Physiol Occup Physiol dynamic contractions is influenced by the stiffness of the tendi- 76(3):253–259 nous structures. J Appl Physiol 99(3):986–994 Kösters A, Wiesinger H, Bojsen-Møller J, Müller E, Seynnes OR Bradley PS, Di Mascio M, Peart D, Olsen P, Sheldon B (2010) High- (2014) Influence of loading rate on patellar tendon mechanical intensity activity profiles of elite soccer players at different per - properties in vivo. Clin Biomech 29(3):323–329 formance levels. J Strength Cond Res 24(9):2343–2351. https :// Kubo K, Kawakami Y, Fukunaga T (1999) Influence of elastic proper - doi.org/10.1519/JSC.0b013 e3181 aeb1b 3 ties of tendon structures on jump performance in humans. J Appl Brownlee TE, Murtagh CF, Naughton RJ, Whitworth-Turner CM, Physiol 87(6):2090–2096 O’Boyle A, Morgans R, Morton JP, Erskine RM, Drust B (2018) Kubo K, Kanehisa H, Kawakami Y, Fukunaga T (2000a) Elastic prop- Isometric maximal voluntary force evaluated using an isomet- erties of muscle–tendon complex in long-distance runners. Eur J ric mid-thigh pull differentiates English Premier League youth Appl Physiol 81(3):181–187 soccer players from a maturity-matched control group. Sci Med Kubo K, Kanehisa H, Kawakami Y, Fukunaga T (2000b) Elasticity Football:1–7 of tendon structures of the lower limbs in sprinters. Acta Physiol Buchheit M, Samozino P, Glynn JA, Michael BS, Al Haddad H, Men- Scand 168(2):327–335 dez-Villanueva A, Morin JB (2014) Mechanical determinants of Kubo K, Ikebukuro T, Yata H, Tsunoda N, Kanehisa H (2010) Time acceleration and maximal sprinting speed in highly trained young course of changes in muscle and tendon properties during strength soccer players. J Sports Sci 32(20):1906–1913 training and detraining. J Strength Cond Res 24(2):322–331 Cavanagh P, Komi P (1979) Electromechanical delay in human skeletal Kubo K, Ikebukuro T, Yata H, Tomita M, Okada M (2011) Morpho- muscle under concentric and eccentric contractions. Eur J Appl logical and mechanical properties of muscle and tendon in highly Physiol Occup Physiol 42(3):159–163 trained sprinters. J Appl Biomech 27(4):336–344 Couppe C, Kongsgaard M, Aagaard P, Hansen P, Bojsen-Moller J, Kubo K, Yata H, Tsunoda N (2013) Effect of gene polymorphisms on Kjaer M, Magnusson SP (2008) Habitual loading results in tendon the mechanical properties of human tendon structures. Springer- hypertrophy and increased stiffness of the human patellar tendon. plus 25(2):343 J Appl Physiol 105(3):805–810 Magnusson SP, Kjaer M (2003) Region-specific differences in Achil- Dowling JJ, Vamos L (1993) Identification of kinetic and temporal les tendon cross-sectional area in runners and non-runners. Eur J factors related to vertical jump performance. J Appl Biomech Appl Physiol 90(5–6):549–553 9:95–95 Malliaras P, Kamal B, Nowell A, Farley T, Dhamu H, Simpson V, Erskine RM, Jones DA, Maganaris CN, Degens H (2009) In vivo spe- Morrissey D, Langberg H, Maffulli N, Reeves ND (2013) Patel- cific tension of the human quadriceps femoris muscle. Eur J Appl lar tendon adaptation in relation to load-intensity and contraction Physiol 106(6):827–838 type. J Biomech 46(11):1893–1899 Finni T, Ikegawa S, Lepola V, Komi P (2003) Comparison of force– Meylan CM, Cronin JB, Oliver JL, Hughes MG, McMaster D (2012) velocity relationships of vastus lateralis muscle in isokinetic The reliability of jump kinematics and kinetics in children of dif- and in stretch-shortening cycle exercises. Acta Physiol Scand ferent maturity status. J Strength Cond Res 26(4):1015–1026 177(4):483–491 Murtagh CF, Vanrenterghem J, O’Boyle A, Morgans R, Drust B, Foster BP, Morse CI, Onambele GL, Williams AG (2014) Human Erskine RM (2017) Unilateral jumps in different directions: COL5A1 rs12722 gene polymorphism and tendon proper- a novel assessment of soccer-associated power? J Sci Med ties in vivo in an asymptomatic population. Eur J Appl Physiol Sport 20(11):1018–1023 114(7):1393–1402 Murtagh CF, Nulty C, Vanrenterghem J, O’Boyle A, Morgans R, Drust Freriks B, Hermens H, Disselhorst-Klug C, Rau G (1999) The rec- B, Erskine RM (2018) The neuromuscular determinants of unilat- ommendations for sensors and sensor placement procedures for eral jump performance in soccer players are direction specific. Int surface electromyography. Seniam 8:13–54 J Sports Physiol Perform 13(5):604–611 Fukashiro S, Besier TF, Barrett R, Cochrane J, Nagano A, Lloyd DG Nagano A, Komura T, Fukashiro S (2004) Effects of the length ratio (2005) Direction control in standing horizontal and vertical jumps. between the contractile element and the series elastic element J Sport Health Sci 2005:272–279 on an explosive muscular performance. Electromyogr Kinesiol Grimshaw P, Fowler N, Lees A, Burden A (2004) BIOS instant notes 14(2):197–203 in sport and exercise biomechanics. Garland Science, New York Onambélé GN, Burgess K, Pearson SJ (2007) Gender-specific in vivo Guadagnoli E, Velicer WF (1988) Relation to sample size to the stabil- measurement of the structural and mechanical properties of the ity of component patterns. Psychol Bull 103(2):265 human patellar tendon. J Orthop Res 25(12):1635–1642 Hansen P, Bojsen-Moller J, Aagaard P, Kjær M, Magnusson SP (2006) Pearson SJ, Onambélé GL (2012) Computation methods affect the Mechanical properties of the human patellar tendon, in vivo. Clin reported values of in vivo human tendon stiffness. J Mech Behav Biomech 21(1):54–58 Mater 5(1):291–297 Helland C, Bojsen-Møller J, Raastad T, Seynnes O, Moltubakk M, Peltonen J, Cronin NJ, Stenroth L, Finni T, Avela J (2012) Achilles Jakobsen V, Visnes H, Bahr R (2013) Mechanical properties of tendon stiffness is unchanged one hour after a marathon. J Exp the patellar tendon in elite volleyball players with and without Biol 215(20):3665–3671 patellar tendinopathy. Br J Sports Med 47(13):862 Reed CC, Iozzo RV (2002) The role of decorin in collagen fibrillogen- Jaric S, Mirkov D, Markovic G (2005) Normalizing physical per- esis and skin homeostasis. Glycoconj J 19(4–5):249–255 formance tests for body size: a proposal for standardization. J Reeves ND, Narici MV (2003) Behavior of human muscle fascicles Strength Cond Res 19(2):467–474 during shortening and lengthening contractions in vivo. J Appl Kastelic J, Palley I, Baer E (1980) A structural mechanical model for Physiol 95(3):1090–1096 tendon crimping. J Biomech 13(10):887–893 Reeves ND, Maganaris CN, Narici MV (2003) Effect of strength train- Kawakami Y, Muraoka T, Ito S, Kanehisa H, Fukunaga T (2002) ing on human patella tendon mechanical properties of older indi- In  vivo muscle fibre behaviour during counter-movement viduals. J Physiol 548(3):971–981 1 3 European Journal of Applied Physiology Reeves ND, Maganaris CN, Ferretti G, Narici MV (2005) Influence of Stenroth L, Cronin NJ, Peltonen J, Korhonen MT, Sipilä S, Finni T 90-day simulated microgravity on human tendon mechanical prop- (2015) Triceps surae muscle–tendon properties in older endur- erties and the effect of resistive countermeasures. J Appl Physiol ance-and sprint-trained athletes. J Appl Physiol 120(1):63–69 98(6):2278–2286 Tillin NA, Jimenez-Reyes P, Pain MT, Folland JP (2010) Neuromus- Reilly T, Thomas V (1976) A motion analysis of work-rate in different cular performance of explosive power athletes versus untrained positional roles in professional football match-play. J Hum Mov individuals. Med Sci Sports Exerc 42(4):781–790 Stud 2(2):87–97 Wang H, Hua C, Cui H, Li Y, Qin H, Han D, Yue J, Liang C, Yang Rosager S, Aagaard P, Dyhre-Poulsen P, Neergaard K, Kjaer M, Mag- R (2013) Measurement of normal patellar ligament and anterior nusson S (2002) Load-displacement properties of the human tri- cruciate ligament by MRI and data analysis. Expert Ther Med ceps surae aponeurosis and tendon in runners and non-runners. 5(3):917–921 Scand J Med Sci Sports 12(2):90–98 Wiesinger H-P, Kösters A, Müller E, Seynnes OR (2015) Effects of Seynnes OR, Erskine RM, Maganaris CN, Longo S, Simoneau EM, increased loading on in  vivo tendon properties: a systematic Grosset J-F, Narici MV (2009) Training-induced changes in struc- review. Med Sci Sports Exerc 47(9):1885 tural and mechanical properties of the patellar tendon are related Wiesinger H-P, Rieder F, Kösters A, Müller E, Seynnes OR (2016) to muscle hypertrophy but not to strength gains. J Appl Physiol Are sport-specific profiles of tendon stiffness and cross-sectional 107(2):523–530 area determined by structural or functional integrity? PLoS One Seynnes O, Koesters A, Gimpl M, Reifberger A, Niederseer D, Nie- 11(6):e0158441 bauer J, Pirich C, Müller E, Narici M (2011) Effect of alpine Wiesinger H-P, Rieder F, Kösters A, Müller E, Seynnes OR (2017) skiing training on tendon mechanical properties in older men and Sport-specific capacity to use elastic energy in the patellar and women. Scand J Med Sci Sports 21(s1):39–46 achilles tendons of elite athletes. Front Physiol 8:132 Stafilidis S, Arampatzis A (2007) Muscle–tendon unit mechanical and Withers R, Maricic Z, Wasilewski S, Kelly L (1982) Match analysis of morphological properties and sprint performance. J Sports Sci Australian professional soccer players. Hum Mov Stud 8:159–176 25(9):1035–1046 Stenroth L, Peltonen J, Cronin NJ, Sipilä S, Finni T (2012) Age-related differences in Achilles tendon properties and triceps surae muscle architecture in vivo. J Appl Physiol 113(10):1537–1544 1 3

Journal

European Journal of Applied PhysiologySpringer Journals

Published: Jun 2, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off