Distal hamstrings tendons mechanical properties at rest and contraction using free‐hand 3‐D ultrasonography

Abstract

Tendon properties impact human locomotion, influencing sports performance, and injury prevention. Hamstrings play a crucial role in sprinting, particularly the biceps femoris long head (BFlh), which is prone to frequent injuries. It remains uncertain if BFlh exhibits distinct mechanical properties compared to other hamstring muscles. This study utilized free‐hand three‐dimensional ultrasound to assess morphological and mechanical properties of distal hamstrings tendons in 15 men. Scans were taken in prone position, with hip and knee extended, at rest and during 20%, 40%, 60%, and 80% of maximal voluntary isometric contraction of the knee flexors. Tendon length, volume, cross‐sectional area (CSA), and anteroposterior (AP) and mediolateral (ML) widths were quantified at three locations. Longitudinal and transverse deformations, stiffness, strain, and stress were estimated. The ST had the greatest tendon strain and the lowest stiffness as well as the highest CSA and AP and ML width strain compared to other tendons. Biceps femoris short head (BFsh) exhibited the least strain, AP and ML deformation. Further, BFlh displayed the highest stiffness and stress, and BFsh had the lowest stress. Additionally, deformation varied by region, with the proximal site showing generally the lowest CSA strain. Distal tendon mechanical properties differed among the hamstring muscles during isometric knee flexions. In contrast to other bi‐articular hamstrings, the BFlh high stiffness and stress may result in greater energy absorption by its muscle fascicles, rather than the distal tendon, during late swing in sprinting. This could partly account for the increased incidence of hamstring injuries in this muscle.

1. INTRODUCTION

The tendinous structures in a muscle‐tendon unit (MTU) are frequently comprised of the aponeurosis and the extra‐muscular free tendon. While the aponeurosis is connected to the muscle fascicle, the free tendons, on the other hand, operate along with the muscle's line of action and are loaded and elongated when muscles produce force. ¹ The free‐tendon interface, where muscle fibers and bones intersect, exhibits distinct mechanical properties that respond to the applied forces. ² This intricate system not only acts as force transmitter to the bones and a shock‐absorbing buffer against sudden impulse to the MTU, but it also plays a pivotal role in the storage and release of elastic energy throughout a wide range of movements. ³ , ⁴ As the MTU elongates, a portion of this elongation is absorbed by the tendons, thereby reducing the necessity for extensive stretching of the muscle fibers. ⁵ This mechanism could be advantageous in minimizing the risk of strain injuries. In the late swing phase, hamstrings tendons, for example, must absorb a considerable amount of energy which is crucial for reducing loads in fatigue conditions. ⁶ , ⁷ This becomes especially critical in fatigued conditions when the muscles are less capable of absorbing energy before reaching the threshold that leads to muscle failure and, consequently, to strain injuries. ⁸ In contrast, stiffer tendons enhance MTU force development and explosiveness which is important for sports performance. This dual function underscores tendons' significance in injury prevention and optimizing performance in sports that require maximum force development.

During sprinting, the tendons of hamstring muscles (semimembranosus [SM], semitendinosus [ST], biceps femoris long head [BFlh], and biceps femoris short head [BFsh]) potentially influence the force regulation. ⁶ , ⁷ Based on the known tendon properties it can be suggested that during the initial stance phase, hamstring tendons with higher stiffness may enhance the overall capacity for force production. ⁶ , ⁷ Conversely, as the late swing phase ensues, tendons exhibiting greater compliance may serve to protect muscle fibers against the risk of strain injuries. ⁶ , ⁷ Previous investigations have established that that hamstrings are a very complex muscle group ⁹ with various architectural characteristics. ¹⁰ However, up to date, very few studies have explored the morphological and mechanical characteristics of hamstrings tendon and, hence, the information for the tendon architecture is scarce. ¹¹ , ¹² , ¹³ Recently, Sahinis et al. ¹¹ has documented significant morphological differences between hamstrings tendons as the ST distal tendon was longer while the SM had greater tendon volume and CSA compared to the remainder hamstrings. In theory, a greater tendon length coupled with a lower CSA may lead to a less stiff ST distal tendon compared to the other hamstring tendons. ⁹ Beyond the variations in morphological characteristics among distal hamstring tendons, noteworthy differences have been identified along the length of each tendon. ¹¹ , ¹² Specifically, research has indicated that the CSA of the hamstring tendon tends to be higher at the midpoint, whereas it diminishes in the distal measurement sites closer to the insertion point. ¹¹ , ¹² Additionally, when considering the mechanical properties of tendons, it has been found that hamstring tendons exhibit greater viscoelasticity properties in comparison to various other tendon structures. ¹⁴ For example, cadaveric ST tendon presents higher elastic modulus (1036 ± 312 MPa) and ultimate stress values (120.1 ± 30.0 MPa) compared to other tendons, including the patellar tendon (417 ± 107 MPa and 76.2 ± 25.1 MPa, respectively). ¹⁴ However, no previous investigations have directly compared individual hamstrings' tendon stiffness combining ultrasound (US) and force measurements.

Two‐dimensional ultrasonography has been used to quantify hamstring muscle‐tendon morphology. ¹⁰ , ¹³ This measurement is limited to a single plane at predetermined measurement locations, which fails to capture the three‐dimensional (3D) shape of muscles and tendons. ¹⁵ Free‐hand 3D US overcomes these limitations as it combines conventional US images with a probe position tracking system that enables the 3D reconstruction of the examined tissue at rest and during tensile loading. Free‐hand 3D US is a reliable tool for the examination of muscle and tendons dimensions in vivo. ¹¹ , ¹³ , ¹⁶

Most of the previous investigations, have explored the mechanical properties of the tendon (mainly of the Achilles and patellar tendon) ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ only under uniaxial loading in the longitudinal direction (i.e., along the line of action of the whole MTU), but this approach provides limiting information regarding tendon function as tendons often change their length and width (transverse strain) simultaneously during muscle contraction. ¹⁹ Multi‐directional strain of the tendinous structures can influence the tendon mechanical properties and muscle function. ¹⁷ , ¹⁹ , ²⁰ Tendon and aponeurosis sheath control muscle bulging by influencing muscles thickness and width changes. Hence, aponeurosis and tendon play a pivotal role in muscle gearing (muscle‐to‐fascicle length change ratio), and the changes in stiffness due to injury, age, or disease, in these structures could impact muscle force. ¹⁹ , ²⁰ These dynamics underscore the critical role of tendon in muscle function potentially contributing to injury susceptibility. ¹⁹ , ²⁰ Furthermore, for relatively long tendons (such as ST), in vitro studies have found variations in elongation along the tendon. ²¹ , ²² , ²³ Collectively, the above studies indicate that tendinous mechanical properties are not uniform ²⁴ and tendon morphological properties can change in both directions (transversally and longitudinally). To the best of our knowledge, the variations in mechanical properties of the hamstrings tendons remain unclear. Multiple‐directional strain, specifically biaxial strain, on tendinous structures is a complex and dynamic aspect of tendon behavior that may have significant implications to understand the role of the tendons.

Tendon stiffness of the hamstrings may influence the muscle fiber strains during dynamic movements, thus altering the risk for injury. ⁶ Even though the role of tendons for hamstring muscle function has been highlighted in recent reviews, ⁴ , ²⁵ there has been no study that have compared the mechanical properties of the individual hamstring tendons. Such information may help explain why BFlh is more frequently injured during sprinting while SM muscle is more frequently injured during long‐stretching. ²⁶ Additionally, if mechanical properties change along each tendon, then this means that susceptibility of the tendons to injury may be region specific. This complexity underscores the importance of understanding how tendons adapt and respond to mechanical loading and the implications for musculoskeletal function and injury prevention. The aim of this study is to examine the mechanical and morphological properties among the hamstrings distal tendons between different measurement locations at rest and during contraction. We postulate that each distinct tendon possesses unique mechanical characteristics, reflecting significant variations in stress and strain responses. Specifically, our expectation is that the distal tendon of ST, owing to its greater length coupled with a smaller CSA, will exhibit reduced stiffness compared to other tendons. Additionally, we hypothesize that the BFlh, with its shorter lever arm and enhanced force‐generating capacity, ⁴ coupled with a stiffer tendon, may display greater stress on its distal tendon. Furthermore, our hypothesis extends to the distal hamstrings tendons as a whole, proposing considerable proximo‐distal differences in mechanical properties along their length when subjected to loading. This suggests that different segments of these tendons undergo diverse mechanical stress and strain levels. Emphasizing the dynamic distribution of loads and mechanical responses along the length of distal hamstrings tendons, the present study may contribute to better understanding of their adaptive mechanisms across various anatomical regions.

2. MATERIALS AND METHODS

2.1. Participants

An a priori power analysis using G*Power (version 3.1.9.2) indicated that, to detect a significant within‐interaction of medium effect size (f = 0.25) with a power of 80% and a correlation coefficient of r = 0.5, a minimum of 11 participants would be required for the repeated measures ANOVA analysis. A total of 15 moderately active (1870 ± 800 MET‐minutes/week) ²⁷ males (mean (SD) age: 22.1 (0.3) years; body mass: 89.1 (3.9) kg; height: 1.86 (0.05) m) volunteered to participate in this study. All participants were fully informed of the procedure, purpose and possible risks and discomfort associated with the study before they gave their written consent to participate. The participants were healthy, and they had no injuries of the lower limbs, including anterior cruciate ligament (ACL) injury or a history of hamstring strain in the 6 months prior to their testing session. The procedures conformed to the Declaration of Helsinki and were approved by the university ethics review committee (ERC‐004/2022).

2.2. Dynamometry

The tests were performed on a Cybex (Humac Norm, CSMI, MA, USA) dynamometer with the subjects in the prone position with the left knee stabilized at 0° (0° = full knee extension) and with the hips in extended position (neutral position) firmly fixed used Velcro straps. If any movements or changes in joint angles were observed during the tests, the trial was excluded, and the procedure was repeated. The dynamometer's rotation axis was carefully aligned with the approximate center of rotation of the knee, on the posterior aspect of the lateral femoral condyle. Axis alignment was evaluated both at rest and during isometric contractions to reduce errors occurring due to joint movement. A twin‐axis goniometer (Model TSD 130B, Biopac Systems, Inc., Goleta, CA, USA) was used to obtain knee angular position. Dynamometer torque and knee joint angular position were obtained simultaneously at 1000 Hz. Additionally, the limb gravitational torque was measured from a static position of 30° of knee flexion to allow correction of the recorded torque for gravitational effects. Following a warm‐up session with submaximal contractions, each subject was asked to perform the maximal voluntary isometric contraction (MVC) of knee flexors for 5–10 s to determine the target contraction levels. This procedure was repeated two times (an additional trial was conducted in cases where the two MVC torque values were not within 5% of each other) with at least 3 min of rest between trials, and the highest value (using a moving average 1‐s epoch) was adopted as 100% MVC and used for further analysis. The participants received verbal encouragement during all MVC tasks. Torque values corresponding to 20%, 40%, 60%, and 80% of participant's MVC were then calculated and the applied force was visually displayed on a screen in real time and was linked to an acquisition system (Biopac Systems Inc., CA, USA). Prior to data collection, subjects underwent a familiarization and a tendon preconditioning protocol involving at least 3–5 consecutive contractions at 80% of MVC or higher, to ensure their ability to linearly increase or decrease moment production at the desired rate during subsequent testing. ¹⁵ After taking 3D freehand US scans from the medial and lateral hamstrings tendons at rest, each subject was asked to perform a sustained submaximal isometric at 20%, 40%, 60%, and 80% of MVC torque (in random order), while transverse 3D freehand US images were obtained during the trials. A rest period of at least 2–3 min was provided between sustained isometric trials to minimize the influence of fatigue.

2.3. Electromyography

Electromyography (EMG) signals were recorded using bipolar bar surface electrodes (inter‐electrode distance 1 cm, TSD 150B, Biopac System Inc., Goleta, CA, USA) in BFlh and ST musculatures. The electrodes were positioned on the BFlh and ST muscle belly where the border of each muscle was carefully identified with an ultrasonic apparatus (SSD‐3500, ALOKA, Japan) according to SENIAM guidelines. Subsequently, the skin was shaved and cleaned with alcohol wipes and the electrodes were attached to the left thigh and fixed with surgical tape. A ground electrode was positioned over a bony landmark on the lateral epicondyle. The EMG signal was amplified (gain ×1000) with an input impedance of 10 MΩ and a common rejection ratio of 130 dB. The EMG produced during the isometric MVC was calculated during a 1‐s epoch at the torque plateau. The signal was filtered using a bandpass filter (low 15 Hz and high 450 Hz, full‐wave rectified) and the root mean square (RMS) was calculated from the raw signal for each trial. EMG activity was expressed as a percentage of maximum EMG during the flexion MVC.

2.4. Freehand 3‐D ultrasound system

US images were obtained with an US apparatus (SSD‐3500, ALOKA, Japan) with a linear array probe of 10 MHz wave frequency, a length of 60‐mm and a frequency of 25 Hz and standardized B‐mode image settings (depth 60 mm, gain 50%, dynamic range 65 dB). ¹¹ , ¹² A molded thermoplastic casing was fixed to the end of the transducer and housed a 2 cm‐thick acoustic standoff pad (Ultra/Phonic Focus, Newark, NJ, USA) to enhance the visualization of scanned tendon CSA. In addition, a six‐degree of freedom Ascension Flock of Birds electromagnetic position sensor (Ascension Technology Corp., USA) was attached to the US transducer and its orientation and position were controlled by a commercial software package (Stradwin [http://mi.eng.cam.ac.uk/rwp/stradwin], v.6.02, Mechanical Engineering, Cambridge University, Cambridge, UK). A video capture card (C039, Avermedia Technologies, Taiwan) was connected to provide live, real‐time US images at a rate of 25 Hz using the video‐capture module of Stradwin software. Prior to scanning, the US transducer and the electromagnetic position sensor were temporally and spatially calibrated in a waterbath (35°C) a single‐wall calibration phantom, similarly to our previous studies. ¹¹ , ¹² Following calibration, pixel coordinates from any acquired 2D US images were transformed into 3D space with an approximate error of ±0.6 mm.

For the US recording, the subjects positioned in the prone stance, with the knee stabilized at 0° (full knee extension) and the hips maintained in an extended position. To standardize the US scanning path during the repeated measurements, first the distal muscle‐tendon unit junctions (MTUJ) were identified and then the fibular head (for BFlh and BFsh), medial tibial condyle (for SM) and pes anserinus (for ST) were recognized and visual echo markers were placed on the skin. A line was drawn on the skin surface with a permanent marker to guide the scan direction commencing from the MTU junction towards the distal ends of each hamstring muscle. Several trial scans for each distal tendon were subsequently taken to better identify the most “appropriate” path of each tendon. This procedure was repeated until three scans with acceptable image quality (i.e., a clear, continuous depiction of the superficial and deep aponeurosis without any sudden changes in the image across the muscle) were obtained. All US measurements were performed by the same examiner, who has experience with musculoskeletal US at least 5 years.

To avoid compression and distortion of the US images, great care was taken during the examination procedure to ensure that consistent minimal pressure was applied with the probe on the examined tissue. Additionally, a generous amount of water‐soluble transmission gel (Transsound, EF Medica Sri, Italy) was applied liberally to the skin, to enhance acoustic coupling and reduce near‐field artifacts. In cases when the distal tendons could not be visualized using a single scan due to an uneven surface, the acquisition was repeated.

2.5. Image analysis

Post‐scanning processing of transverse US images was performed using Stradwin software (Version 6.0). First, distal hamstrings tendon boundaries (CSAs) were manually segmented on each image along the length of each tendon at ~5 to 10‐mm intervals. ¹¹ , ¹² , ¹⁶ Subsequently, surface interpolation through the segmentation contours created a rendered 3D image of the distal tendon from each scan ¹¹ , ¹² and individual tendon volume reconstructions were conducted on the digitized cross sections using the volumetric algorithms provided in Stradwin ¹¹ , ¹² (Figures 1 and 2). Then, the distal tendon length (mm) was quantified as the distance between the distal end of the most distal fascicle (without any muscle fiber attachment) and the insertion of its tendon according to a procedure similar to our previous studies. ¹¹ , ¹² In summary, for each tendon length landmarks were then positioned roughly at the center of each transverse slice. Throughout this process, the placement of each landmark was visualized using a combination of frontal, sagittal, and transverse image reslices to ensure optimal positioning near the center of the examined structure across all reconstructed slices. Following this, the landmark coordinates were exported to Excel (Microsoft Office Excel 2016, Microsoft, Redmond, WA, USA) and fitted using cubic splines. Subsequently, the distances between successive landmarks were measured and cumulatively added to determine the overall tendon length. ¹¹ , ¹² Then, the anatomical CSA (mm²) as well as the widths (anteroposterior diameter [APD] and mediolateral diameter [MLD]) were calculated from three measurement sites along the distal tendon: at 20% (Proximal), 50% (Middle), and at 80% (Distal). For the measurement of anatomical CSA, a region of interest (ROI) was defined manually as the tendon region without bone or fascia with the best reflection.

FIGURE 1.

Examples of three‐dimensional ultrasound reconstruction, and cross‐sectional area at three different measurement sites along the tendon of the semitendinosus (ST; A, B) and semimembranous (SM; C, D) muscle at rest (A, C) and during contraction (B, D) of the 80% of maximal torque. Three‐dimensional ultrasound image reconstructions were performed in Stradwin software using manually digitized cross sections of the tendon. Note that the tendon length was increased during the contraction but the tendon volume was remained unaltered.

FIGURE 2.

Examples of three‐dimensional ultrasound reconstruction, and cross‐sectional area at three different measurement sites along the tendon of the biceps femoris long head (BFlh; A, B) and biceps femoris short head (BFsh; C, D) muscle at rest (A, C), and during contraction (B, D) of the 80% of maximal torque. Three‐dimensional ultrasound image reconstructions were performed in Stradwin software using manually digitized cross sections of the tendon. Note that the tendon length was increased during the contraction but the tendon volume was remained unaltered.

Distal tendon elongation was quantified by subtracting mean length obtained at each corresponding MVC level from the resting length. Tendon longitudinal strain was calculated by dividing tendon displacement by the corresponding length at rest [(Length_MVC%−Length_rest)/Length_rest]. CSA, MLD and APD strains were determined by dividing the difference in diameter by the respective measurements at rest. Tendon stress was quantified by dividing tendon force by the minimal CSA (i.e., stress MPa) for each tendon. ²⁸ , ²⁹ Tendon force (F) was estimated by dividing knee flexion torque (M) and moment arm (r) of each hamstring muscle (F = M/r). To account for the contribution of multiple muscles in generating knee flexion torque, we adjusted each muscle's contribution based on its physiological cross‐sectional area (PCSA) relative to the total hamstring muscle group CSA. ³⁰ The tendon moment arm was estimated using regression equations as a function of the knee joint angle based on equations which were presented by Herzog and Read. ³¹ The individual hamstrings tendon stiffness was quantified by the slope of the line that fits the tendon force‐elongation curve, specifically at 20%, 40%, 60%, and 80% of MVC force and the corresponding tendon elongation, as has been previously described. ³² , ³³

2.6. Statistical analysis

Statistical analyses were performed using SPSS (version 25.0, IBM, Armonk, NY, USA). The normality of the data was assessed using the Kolmogorov–Smirnov test. A three‐way ANOVA (4 × 3 × 4) was applied to examine CSA, AP and ML strain between the hamstring's tendons, at the different measurement sites at the different torque levels. The differences in tendon elongation (in absolute and relative values), tendon stress and volume changes between hamstrings and torque levels were examined using two‐way ANOVA (4 × 4). Further, one way ANOVA with repeated measures was used to explore the differences in individual tendon stiffness. Significant interactions were followed up with simple effects tests and, if significant, post‐hoc Tukey tests were applied to investigate significant differences between pairs of means. Statistical significance was set at p < 0.05.

3. RESULTS

3.1. Tendon displacement

Of the four distal hamstrings tendons, ST presented the largest while the BFsh the smallest tendon length at rest (p < 0.01). No further differences were observed (p > 0.05) (Table 1).

TABLE 1.

Hamstrings distal tendon morphological characteristics at rest across the different measurement sites.

	Semitendinosus	Semimebranosus	Biceps femoris LH	Biceps femoris SH
Length (mm)	132.43 ± 14.44*	56.76 ± 3.54	62.37 ± 0.93	32.12 ± 5.17**
Volume (mL)	1.29 ± 0.19	2.97 ± 0.40	2.34 ± 0.22	2.03 ± 0.18
Proximal site
CSA (mm2)	17.59 ± 3.5**	96.08 ± 2.48***	61.48 ± 3.66	95.59 ± 2.99***
AP diameter (mm)	2.17 ± 0.12**	4.20 ± 0.85	3.51 ± 0.3	6.51 ± 0.59*
ML diameter (mm)	8.34 ± 0.4**	22.6 ± 1.76*	21.93 ± 0.67	17.36 ± 1.54
Middle site
CSA (mm2)	27.57 ± 1.12	118.95 ± 7.64***	81.62 ± 4.35	96.72 ± 5.39***
AP diameter (mm)	3.28 ± 0.31**	4.15 ± 0.2	4.21 ± 0.37	5.3 ± 0.59*
ML diameter (mm)	8.31 ± 0.89**	26.07 ± 2.13*	22.45 ± 0.88	15.75 ± 2.17
Distal site
CSA (mm2)	21.42 ± 0.71	49.72 ± 3.13	49.29 ± 1.71	46.86 ± 1.98
AP diameter (mm)	2.55 ± 0.14	2.57 ± 0.4	3.71 ± 0.21	2.56 ± 0.26
ML diameter (mm)	8.76 ± 0.33**	14.63 ± 0.45	14.37 ± 0.95	15.7 ± 1.56

Abbreviations: AP, Anterior–posterior; CSA, Cross‐sectional area; LH, Long Head; ML, Mediolateral; SH, Short Head.

^*Greater compared to the other tendons (p < 0.05).

^**Lower compared to the other tendons (p < 0.05).

^***Greater compared to the ST and BFlh (p < 0.05).

Absolute tendon displacement and the strain of the different hamstring tendons at various isometric contraction intensities are presented in Data S1 and Figure 3, accordingly. The ANOVA indicated a significant tendon length × torque level interaction (F _9,126 = 54.04, p < 0.0001), with the ST exhibiting the most substantial displacement from the lowest to the greatest level of force, surpassing the other tendons. Following, the SM and BFlh showed comparable length changes, with no significant differences between them. Notably, BFsh displayed the smallest alteration in tendon length. Additionally, for all the distal hamstrings tendons, greater change was observed at the higher contraction intensities compared to the lower ones. For tendon strain, no significant tendon × torque level interaction (F _9,126 = 1.60, p > 0.05) was found. However, there was a significant main effect of tendon (F _3,42 = 20.99, p < 0.001) on strain with ST strain being greater than the other tendons' strain (p < 0.001). Tendon strain increased linearly with increasing target torque levels for all tissues (F _3,42 = 462.27, p < 0.001).

FIGURE 3.

Tendon displacement of the semitendinosus (A), semimembranosus (B), biceps femoris long head (C), and biceps femoris short head (D) across the four submaximal target torque levels (20%, 40%, 60%, and 80% MVC torque). The black line indicates the mean and each data point corresponds to one participant. * Greater compared to other tendons (p < 0.05).

Individual tendon stiffness values are shown in Data S2. The one‐way ANOVA with repeated measures indicated a significant main effect of tendon (F _3,42 = 95.57, p < 0.0001) with the ST (43.87 ± 7.60 N/mm⁻¹) presented the lowest stiffness while the BFlh (730.46 ± 87.86 N/mm⁻¹) presented the greatest. No differences we observed between the SM (438.25 ± 76.02 N/mm⁻¹) and BFsh (413.59 ± 91.44 N/mm⁻¹) tendon stiffness.

3.2. Cross‐sectional area strain

Hamstrings distal tendons CSA values at rest are presented in Table 1. The ANOVA yielded a significant tendon × measurement site interaction (F _6,84 = 231.11, p < 0.001). Of the four tendons, the SM and BFsh showed the greatest CSA and the ST the lowest (p < 0.05). The middle measurement site presented the greatest CSA while the lowest was observed distally (p < 0.001).

The mean group values of the CSA strain are displayed in Figure 4. The ANOVA revealed a significant three‐way interaction (tendon × measurement site × torque level) (F _18,252 = 14.05, p < 0.0001). Post hoc Tukey tests indicated that ST presented the highest across all the torque levels and measurement sites while SM the lowest CSA strain especially the middle and distal sites in higher target forces (60% and 80% MVC force). No differences were observed between BFlh and BFsh tendons. Additionally, CSA at the proximal site presented the lowest strain in ST and SM in most target forces, while CSA at the middle site presented the lowest strain compared to the other sections in the BFlh and BFsh. Finally, the greatest strain was observed at 80%, followed by 60%, 40%, and 20% of MVC.

FIGURE 4.

Tendon cross‐sectional strain of the semitendinosus (A), semimembranosus (B), biceps femoris long head (C), and biceps femoris short head (D) at the proximal (open circles), middle (light gray circles), and distal (dark gray circles) measurement sites across the four submaximal target torque levels (20%, 40%, 60%, and 80% MVC torque). The black thick line indicates the mean and each data point corresponds to one participant. * Greater compared to other tendons (p < 0.05), ^ lower compared to other tendons (p < 0.05).

3.3. Anteroposterior and mediolateral diameter strain

At rest, there was a significant tendon × measurement site interaction for the AP (F _6,84 = 84.22, p < 0.001) and ML (F _6,84 = 143.11, p < 0.001) diameter (Table 1). Between hamstring tendons, BFsh presented the greatest and the ST the lowest AP diameter, while the SM presented the greatest and the ST the lowest ML diameter (p < 0.01). The greater AP and ML diameters were observed at the middle measurement locations, while the lowest was distally (p < 0.01) (Table 1).

The mean AP and ML diameters strain of distal hamstring tendons are shown in Figures 5 and 6, respectively. A significant tendon × measurement site × torque level interaction was identified for both the AP (F _18,252 = 7.64, p < 0.001) and ML (F _18,252 = 14.46, p < 0.0001) diameter strain. Post hoc analysis reveal that the ST presented the greatest followed by SM and BFlh whereas the BFsh presented the lower AP diameter strain (p < 0.01). Also, significant differences were observed between proximal and middle with the distal measurement site (p < 0.05) in AP diameter. Additionally, the greater ML diameter strain presented by ST and BFlh followed by SM while the BFsh presents the smaller strain. Regarding the measurement location, the middle site presented the lowest ML diameter strain while no further differences were observed. Greater strain for both the AP and ML diameter were observed at the higher contraction intensities compared to the lowest ones.

FIGURE 5.

Tendon anteroposterior strain of the semitendinosus (A), semimembranosus (B), biceps femoris long head (C) and biceps femoris short head (D) at the proximal (open circles), middle (light gray circles), and distal (dark gray circles) measurement sites across the four submaximal target torque levels (20%, 40%, 60%, and 80% MVC torque). The black thick line indicates the mean and each data point corresponds to one participant. * Greater compared to other tendons (p < 0.05), ^ lower compared to other tendons (p < 0.05).

FIGURE 6.

Tendon mediolateral strain of the semitendinosus (A), semimembranosus (B), biceps femoris long head (C), and biceps femoris short head (D) at the proximal (open circles), middle (light gray circles), and distal (dark gray circles) measurement sites across the four submaximal target torque levels (20%, 40%, 60%, and 80% MVC torque). The black thick line indicates the mean and each data point corresponds to one participant. * Greater compared to other tendons (p < 0.05), ^ lower compared to other tendons (p < 0.05).

3.4. Stress

Figure 7 shows the mean stress of the distal hamstrings tendons across the torque levels. The two‐way ANOVA revealed a significant tendon × torque level interaction (F _9,126 = 259.591, p < 0.001) on hamstrings tendon stress. The BFlh presented the greatest stress, followed by SM and ST while the BFsh presented the lowest one (p < 0.001). Tendon stress increased linearly with the level of torque (p < 0.001).

FIGURE 7.

Tendon stress of the semitendinosus (A), semimembranosus (B), biceps femoris long head (C), and biceps femoris short head (D) across the four submaximal target torque levels (20%, 40%, 60%, and 80% MVC torque). The black line indicates the mean and each data point corresponds to one participant. $ Greater compared to BFsh (p < 0.05), * greater compared to ST and BFsh (p < 0.05) and, # greater compared to other tendons (p < 0.05).

3.5. Tendon volume

The ANOVA yielded a significant effect on tendon (F _3,42 = 106.54, p < 0.001), with the SM displaying the greatest volume compared to the other tendons, followed by BFlh and BFsh. Nevertheless, the ST showed the smaller distal tendon volume among the tendons investigated (p < 0.01) (Table 1).

The mean group values of the tendon volume changes at various torque levels are shown in Data S3. The ANOVA did not find any significant tendon × torque level (F _9,126 = 0.90, p > 0.05) interaction or main effect of tendon (F _3,42 = 0.89, p > 0.05) or target torque level (F _3,42 = 0.17, p > 0.05) on tendon volume.

3.6. Muscle activation

For normalized EMG amplitudes of ST and BFlh, the two‐way ANOVA revealed a non‐significant muscle × torque level interaction (F _3,42 = 0.21, p > 0.05) or main effect of muscle (F _1,14 = 0.3, p > 0.05) (Data S4). However, a significant main effect of torque level was found (F _3,42 = 10.71, p < 0.0001) with the EMG of both muscles being higher as the level of torque increased.

4. DISCUSSION

The major finding of this study is that mechanical properties, such as stress and strain, differ between the four distal tendons of the hamstrings. In particular the ST presented the greatest strain and the lowest stiffness while the BFlh presenting the greatest stress and stiffness among the hamstrings distal tendons which confirm our initial hypotheses. Furthermore, proximo‐distal differences in distal tendons' mechanical properties were identified with the proximal measurement site presented lower CSA strain and APD strain compared to the other measurement sites.

4.1. Tendon displacement

The results of this study indicate significant differences in length between hamstrings distal tendons which is consistent with previous investigations. ¹¹ , ¹² Further, significant differences were observed in tendon displacement between hamstrings at different torque levels (Data S1, Figure 3). In particular, the ST had the greatest change, followed by SM and BFlh while the BFsh presented the lowest. Additionally, the ST showed the lowest stiffness (higher compliance) while the BFlh presenting the greatest stiffness among the four distal tendons (Data S2). This is in contrast to Kellis et al. ³⁴ who observed that the distal tendon‐aponeurosis of the BFlh undergoes greater strain than that of the ST when the knee is extended passively to 90°. The discrepancies between the two studies may be due to different testing conditions (isometric contractions vs. passive movement). Further, Kellis et al. ³⁴ only measured strain of a single segment of the aponeurosis which may not provide a full representation of the overall tendon or MTU properties. Different mechanical properties between tendons may be due to differences in CSA and in most cases, are related to the distinct properties of their connected muscle bellies. For example, as a result of its high mass and relatively short muscle fiber length couple with short tendon, SM has the greatest force‐generating capacity compared to ST which has longer fascicles, greater tendon length, and lower mass. ⁴ Based on these results, it seems that the ST could be considered a more compliant tendon, while the distal tendons of SM, BFlh, and BFsh as stiffer tendons as they presented smaller changes in tendon length during the different target forces. From a mechanical standpoint, compared to a complaint tendon, stiffer tendons (such as SM, BFlh, and BFsh) may experience significant less strain at a given load, have greater potential to produce greater force and store more energy at a given strain. ³⁵

The effect of tendon compliance in injury risk is not clear. From the above findings, it seems that the ST may be less prone to injury risk due to its higher tendon compliance, which could act as a mechanical buffer that reduces muscle fibers stretch and protects against injury. ³⁶ In particular, ST coupled with a distal compliant tendon, and given its parallel fiber arrangement, may lower the risk of muscle strain injuries. ⁵ , ⁸ , ³⁷ On the contrary, the BF connected to a stiffer distal tendon and characterized by its pennated fiber orientation, is more susceptible to strain injuries, especially when subjected to high force and multiaxial loading conditions. ⁵ , ⁸ , ³⁷ However, others have suggested that a reduction in tendon compliance could contribute to an increase in injury risk during sprinting, unless the changes in running posture enable the musculotendon complex to function at a shorter length. ⁶ Despite the similar morphological characteristics and tendon properties with the SM, the BFlh is more susceptible to injury due to greater changes in MTU length during sprinting. ⁶ This could be attributed to the larger moment arm in the hip joint and the smaller moment arm in the knee joint observed in BFlh, which collectively increases the risk of injury compared to the SM. ⁴ The stiffness of tendons plays a crucial role in shaping both the force‐length and force‐velocity profiles of the muscles. ¹ , ³⁸ Concerning the force‐length relationship, tendons with higher compliance could broaden the operative range by shifting the optimal length of the muscle towards greater lengths. ¹ Moreover, compliant tendons (such as distal ST tendon), could also reduce muscle work requirements and increase efficiency by allowing the muscle to contract over a shorter length range during locomotion. ³⁸ In terms of the force‐velocity relationship, compliant tendons can enhance overall MTU work and efficiency per cycle by increasing muscle force and decreasing muscle shortening velocity during contraction cycles. ³⁸ A higher compliance will allow muscle fibers to contract at lower shortening velocities than the whole MTU, and as a consequence of the force–velocity relationship, their force‐generating potential will be higher. ¹ , ³ For example, during sprinting the compliant distal tendon of ST could enable the muscle fibers to contract at a lower velocity, enhancing the force generation during this movement. This increased force potential can contribute to better performance and reduced risk of strain injuries, especially in activities that involve high force production. Nevertheless, if a tendon is too compliant, the muscle should shorten rapidly in order to stretch the tendon and store elastic energy prior to the onset of MTU shortening, which raises metabolic cost and lowers efficiency. ⁷ More research is needed to understand the interactions between hamstring muscles and tendons, as well as the effects of tendon properties on hamstrings function in vivo. Future research should explore how changes in tendon properties induced by training or injury affect muscle function and economy in different types of movement.

4.2. Tendon cross‐sectional area strain

The augmented CSA strain presented by ST relative to the other tendons (Figure 4) could be beneficial as it increases the ability to store and release mechanical energy and enables energy dissipation when it is required. ¹ This is line with the whole MTU function, as Magnusson et al. ³⁹ observed greater strain of ST MTU compared to BFlh during a passive stretching maneuver. It was also interesting that the proximal site of the distal hamstrings tendons underwent generally smaller CSA strain compared to other measurement sites (Figure 4). To our knowledge, this is the first study that examines the CSA strain between different locations in hamstrings tendons. The nonuniformity of (axial) strain along its length has been previously documented in ST. ²² The non‐uniform behavior of CSA along the tendons length could be attributed to their insertion. ⁴⁰ Specifically, tendons with wider insertion point may deviate from behaving like a single linear tensile element, resulting in non‐uniform mechanical responses during muscle contraction. ⁴¹ At the knee level, anatomical studies have documented the attachment points of tendons within the hamstring group exhibit great variability. ⁹ Specifically, the distal free tendon of the ST wraps around the tibia, forming multiple bands before integrating into the pes anserinus. ⁴² Further, the distal tendon of the SM also displays arms or bands, some interweaving with branches of the posterior oblique ligament. ⁴³ This configuration is thought to enhance the SM role in stabilizing rotation and controlling knee hyperextension. ⁴³ Similarly, in animal studies variation in tendon material properties was observed in synergistic muscles, indicating that tendon is a dynamic tissue with the potential to modulate material properties in response to mechanical demand. ⁴⁴ In summary, these findings indicate that for a given tensile force, the ST may experience greater CSA strain and shear loading than other tendons, while the SM the lowest. Also, the middle and distal portions of the tendons may experience greater shear loading than the proximal especially in ST and SM tendons due to the greater CSA strain observed at these tendon locations. This distinct loading pattern emphasizes the intricate regional variations in how tendons respond to mechanical forces, advancing our knowledge of how tendons adapt and function under different mechanical conditions.

4.3. Anterior–posterior and mediolateral diameter strain

Semimembranosus presented the greatest APD and MLD at rest, whereas the ST the lowest. In line with previous investigations on the free Achilles tendon in vivo, ¹⁶ , ¹⁷ hamstrings tendons elongated longitudinally while narrowing transversely during contraction at different torque levels. The greatest APD and MLD strain were identified for ST, while the BFlh and BFsh presented the lowest one (Figures 5 and 6). Additionally, significant differences were observed along the tendon length (Figures 5 and 6). This variation may suggest internal reorganization of free tendon fascicles, which might be essential to minimize areas of stress and strain concentration during muscle contraction. ⁴⁵ Also, it may be related to tendon functional properties, such as providing a transition between compliant muscle and stiff bone. ⁴⁶ The approximation of a tendon as a single, uniform structure may hinder our understanding of the processes that regulate tendon adaptations and maladaptations. ¹⁵ Studying tendons as uniform entities, could overlook the intricate and dynamic nature of its composition and mechanical properties. ¹⁵ This oversimplification impedes our ability to unravel the complexities underlying tendon responses to mechanical loading, pathology, and exercise rehabilitation. The assessment of the APD and MLD strain provides a comprehensive analysis of tendon deformation during contraction, offering insights into both length and width changes. This combined approach enables to investigate the holistic response of the hamstrings tendon to mechanical forces, capturing the intricate dynamics of their biomechanical properties.

4.4. Stress

The present finding indicated significant differences in tendon stress among hamstrings tendons, with the BFlh presenting the greatest values, whereas the SM the lowest values. The present results are not surprising as the BFlh exhibits a shorter lever arm at the knee joint and along with architectural features that enable to generate higher level of force. ⁴ This augmented torque production is combined with a stiffer tendon, which reduces the tendon strain during force generation. Consequently, the interplay of a shorter lever arm, enhanced force‐generating capacity, and tendon stiffness could increase the stress placed on the distal tendon. Even though BFlh and SM display comparable architectural properties ⁴ , ⁹ the fact that SM has greater tendon CSA and greater moment arm compared to the BFlh could explain the reduced stress placed on its tendon.

Variation in tendon stress was also observed along the tendon length which is consistent with previous investigations in Achilles tendon. ¹⁸ Reeves & Cooper ¹⁸ found lower tendon stress at the proximal compared to the distal measurement location. A more compliant tendon (such as distal ST tendon) could undergo greater longitudinal strain and average stress under the same contraction‐induced force and hence might have a greater possibility of both stress and strain induced injury compared to a stiffer tendon with the same morphological characteristics. ³⁸ Hamstrings tendons deformation is region specific and closely related to differences in mechanical properties between the proximal, middle and distal components of hamstrings tendons.

5. PERSPECTIVES

The study reveals crucial insights into the mechanical properties of distal hamstring tendons. The identification of the ST as a more compliant tendon, experiencing greater strain, implies its potential role in energy storage and dissipation. Additionally, proximo‐distal differences in mechanical properties highlight the importance of location within the tendon, contributing to a nuanced understanding of hamstring function. For example, by examining the impact of grafts for ACL reconstruction surgery on the mechanical dynamics of distal hamstring tendons, could shed light on potential alterations in the overall functional integrity of these tendons post‐surgery. The findings contribute to the broader discourse on tendon mechanics, urging further exploration of how training‐ or injury‐induced modifications in tendon properties impact overall muscle function and efficiency. In particular, delving into the potential correlation between distal hamstring tendon mechanical properties and hamstring injuries, as well as sprint performance, provide a valuable perspective. Understanding these relationships is crucial for developing targeted rehabilitation strategies and optimizing athletic performance. The study underscores the need for a comprehensive understanding of the dynamic interplay between muscles and tendons, especially in diverse physiological and mechanical conditions.

6. LIMITATIONS

There are several limitations to the current study that should be taken into account. Stiffness values which are reported in this study were obtained during submaximal efforts and, thus, they correspond to midrange of stress experienced by the distal hamstrings tendon in maximal efforts. Although a higher force level would have certainly induced higher tendon stiffness, it could be more difficult to maintain a high force contraction level throughout the scanning process, which would influence the consistency and accuracy of our measurements. The generalizability of findings to real‐life movements remains uncertain, primarily due to the dynamic nature of tendons. ¹ Factors such as joint angles and movement speed play pivotal roles in influencing tendon behavior, ¹ making it challenging to extrapolate results obtained under controlled conditions (isometric contractions) to real‐world scenarios (such as sprinting). Additionally, maintaining a high force contraction level may result in tendon creep, which would cause a reduction in tendon stiffness. Another limitation of the study is that we only tested the 0° knee angle due to the difficulty to assess tendon properties at other knee angles. This limitation may affect the generalizability of the results to different knee angles or hip angles. Since the study focused exclusively on a single knee angle, it does not account for potential variations in tendon behavior in other knee angles, which could be critical for a comprehensive understanding of the dynamic changes in tendon properties during muscle contraction. Although hamstring tendon slack length might impact our findings, we did not directly measure it. Considering existing inconsistencies in the literature, we were cautious about introducing bias by relying on a specific slack length estimate, which could result in substantial changes and erroneous inferences about muscle function. ⁹ Despite previous studies highlighting high test–retest reliability for distal hamstrings tendon morphological properties at rest, ¹¹ , ¹³ there is a notable gap in our understanding of the reliability of the mechanical properties of distal hamstring tendons, which warrants further exploration. There are different ways to estimate the moment arm depending on the equipment and methodologies used, but in general, it requires precise measurements of the muscle and joint positions and the associated kinematic and kinetic data. In the present study, hamstrings moment arms were estimated from regression equations provided by previous data ³¹ and, hence, they may contain errors. The use of PCSA to estimate muscle contributions to knee torque has limitations. It assumes uniform muscle activation, not capturing the dynamic complexity of muscle interactions. Additionally, our model assumes negligible antagonistic muscle activity, potentially leading to underestimates in hamstrings force due to dynamic physiological interactions. Finally, a typical limitation of models, such as the one which was used in this study, is that tendon stiffness values were estimated under the assumption that the force of each muscle, which was calculated based on estimated values of the moment‐arm, was considered proportional to its PCSA. This may introduce errors in the tendon force–elongation relationship.

7. CONCLUSION

This study reveals variations in distal hamstring tendons dimensions at rest. The ST distal tendon presented the longest length and the lower CSA, while the SM exhibited the greatest tendon volume and CSA compared to other distal tendons. Among the distal hamstring's tendons, the ST presented the greatest strain (lower stiffness), indicating its capacity to stretch or deform under load more compared to other tendons. In contrast, BFlh distal tendon showed the greatest stiffness and stress, suggesting that it could experience greater tension when the hamstring muscle group contracts. Compared to other bi‐articular hamstrings, the high stress and low strain (greater stiffness) of the BFlh may lead to more energy absorption by its muscle fascicles than the distal tendon during late swing in sprinting. This may contribute to the heightened risk of hamstring injuries in this muscle. Further, the mechanical properties change along each hamstring distal tendon, having an influence on the mechanics of the tendons themselves. It is concluded that explanation of mechanical behavior of each individual hamstring during various movements should consider the mechanical properties of their tendons.

FUNDING INFORMATION

No funding was received for this research.

CONFLICT OF INTEREST STATEMENT

The authors declare that there is no conflict of interest regarding the publication of this paper. No competing financial interests or personal relationships that could have influenced the work reported in this paper are present.

ETHICS STATEMENT

This study was conducted in accordance with the ethical standards of Aristotle University ethics committee and received approval (ERC‐004/2022). All procedures involving human participants were performed following ethical guidelines, ensuring confidentiality, informed consent, and compliance with relevant regulations.

CONSENT

Informed consent was obtained from all individual participants included in the study. Participants were provided with comprehensive information about the study objectives, procedures, potential risks, and benefits before obtaining their consent.

Supporting information

Data S1:

Data S2:

Data S3:

Data S4:

Sahinis C, Kellis E. Distal hamstrings tendons mechanical properties at rest and contraction using free‐hand 3‐D ultrasonography. Scand J Med Sci Sports. 2024;34:e14621. doi: 10.1111/sms.14621

DATA AVAILABILITY STATEMENT

The data supporting the findings of this study are available upon reasonable request from the corresponding author.