Tendons are specialized musculoskeletal structures responsible for transferring forces between muscles and bones. The frequency, duration, and/or magnitude of tendon forces can change dramatically in response to changes in physical activity and muscle strength. It seems reasonable to expect that if muscle tendon loading changes, then the tendon's structural properties will change so that joint mechanics are maintained (e.g., the joint angle at which peak torque is achieved remains the same), and tendon strains are limited to prevent tendon injury. Surprisingly, there are few published data describing the temporal adaptation of tendon to altered mechanical loading, particularly in response to muscle strength training.
Numerous studies have quantified tendon mechanical properties (1-3,5,14,15,17,19-21,25,27) and how these properties are affected by age (11), gender (12), disuse (27), overload (28), endurance training (6-8,24,26), and resistance training (10,11,13,20,23). However, few studies have quantified the temporal adaptation of tendon in response to increases in muscle strength. Kubo et al. (13) have shown a 19% increase in muscle strength, after 8 wk of strength training, accompanied by an increase in tendon stiffness, an increase in passive muscle tendon stiffness, and no change in the deformation induced in the tendon during a maximum isometric muscle contraction. Two studies involving 6 months of resistance exercise training in elderly subjects found the stiffness of tendon-aponeurosis structures decreased while the stiffness increased in a control group (11,22). Tendon mechanical properties have been shown to change with strength training, but it is not clear how quickly these changes occur or how these changes affect tendon deformation and strains during physiological loading.
The maximum deformation and strain induced in tendons from in vivo muscle loading are known to vary considerably. Ultrasound studies performed in human adults (4,10,12,13,15,17-19,21,24), the elderly (11,23), and in female and male individuals (12) have reported average in vivo maximum strain values ranging from 2.5% (17) to more than 10% (10-12,23). Clearly, this large variation in maximum strain values suggests large variations in muscle tendon structural properties and joint mechanics. It is not clear whether these differences represent individual preferred strain limits or, perhaps, differences in specific muscle- and tendon-adaptation responses.
The objectives of this investigation were 1) to further characterize the variations in tendon strains experienced in vivo, 2) to quantify the interaction between muscle strength and tendon strain through the course of a strength training program, and 3) to test the hypothesis that tendons adapt to changes in muscle strength to maintain strains within a preferred operating range. To achieve these objectives, human subjects performed an 8-wk strength training program for the ankle plantar flexor muscles (i.e., triceps surae muscle). Ultrasonography and joint testing procedures were used to quantify, at regular intervals throughout the strength training program, changes in triceps surae muscle strength and peak strains induced in the Achilles tendon during maximum isometric plantar flexion efforts.
The right lower leg of 10 male subjects, age 24.9 ± 3.4 yr (mean ± SD), mass 78.1 ± 9.7 kg, and height 176.5 ± 7.2 cm, were tested throughout an 8-wk ankle plantar flexion strength training program. All subjects were healthy, moderately active, and had no history of lower-extremity musculoskeletal injuries. Subjects were surveyed and excluded from the study if they participated in any regular strength or endurance activities, defined as participating in a strength and/or endurance training program more than twice per week for more than 1 month, during the year before they were surveyed. All procedures were approved by the University of California-Davis Medical Center human subjects institutional review board. All subjects were informed about the study and gave written consent to participate.
Subjects performed an 8-wk strength-training program consisting of three weekly sessions separated by at least 1 d of rest. Subjects trained using a machine that allowed them to sit and perform heel-raises with a variable weight applied to the top of their thighs. The strength training sessions consisted of three sets of ten heel-raising lifts involving 70% of the maximum force the subject generated during the previous testing session.
Subjects were tested before and at the end of the first, second, fourth, sixth, and eighth weeks of the strength training program. To ensure consistency, the same investigator supervised the training and acquired all measurements throughout the study. Body weight and height were measured at the start of the study. Subjects performed a brief warm-up consisting of a 5-min jog and basic stretching exercises before each testing session. Pilot studies demonstrated that 5 min of cyclic loading of the Achilles tendon allowed any dynamic creep response to stabilize and the strain induced during subsequent cyclic loading to vary by less than 0.25%. Subjects knelt with their right leg on a custom bench (Fig. 1) that incorporated a foot stirrup and force-transducer (Omega Model LCCA 500, 2224 N capacity) assembly to quantify plantar flexion force (FPF). Calibration of the force-transduction system yielded a correlation coefficient of 1.0. The resolution, precision, and accuracy of the system for quantifying plantar flexion force were found to be 0.6 N, 0.1%, and better than 0.6 N, respectively. The bench prevented movement of the lower leg in both the vertical and horizontal directions and kept the knee flexed at 90°. The vertical location of the force transducer and stirrup were adjusted for each subject to locate the foot stirrup at the metatarsophalangeal joints and to maintain the right ankle at an angle of 90°. Achilles tendon lever arm ratio (LAR), defined as the perpendicular distance from the lateral malleolus to the line of action of the force-transducer cable divided by the perpendicular distance from the lateral malleolus to the line of action of the Achilles tendon, was determined using calipers. The line of action of the Achilles tendon was defined by the lateral aspect of the midline of the Achilles tendon. The midline was determined by palpation and visual inspection. LAR was used to calculate the Achilles tendon force (FAT) from the plantar flexion force (FAT = FPF × LAR). Because the Achilles tendon acts in series with the triceps surae muscle, triceps surae muscle force was assumed equal to the Achilles tendon force.
A Hitachi EUB 405 PLUS Ultrasound system with a 64-mm, 7.5-MHz linear probe was used to obtain images of the Achilles tendon, defined as the region between the calcaneus bone-Achilles tendon junction and the Achilles tendon-soleus muscle junction, throughout the testing. The 7.5-MHz frequency was the maximum frequency available from the Hitachi EUB 405 system and provided the greatest spatial resolution for that system. The Achilles tendons examined in this study ranged in length between 56.82 and 90.80 mm. The 64-mm-long linear ultrasound probe was not able to image the entire Achilles tendon at one time; therefore, several steps were required to obtain images that could be used to accurately quantify Achilles tendon length throughout a plantar flexion effort. First, a hollow 1-mm-diameter plastic rod was secured to the subject's skin, with tape and spray adhesive, midway between the calcaneus bone and the soleus muscle-Achilles tendon junction. Second, the ultrasound probe was oriented in a sagittal plane and located along the dorsal aspect of the Achilles tendon, in a position that provided either an image of both the plastic rod and the soleus-Achilles tendon junction or an image of the plastic rod and the Achilles tendon-calcaneus junction. Conducting gel was placed between the probe and the skin. The investigator rested his hand lightly on the subject's skin and placed the probe in his hand, to maintain the probe parallel to the Achilles tendon. Third, subjects performed six repeat ankle plantar flexion trials (described in the next paragraph). The probe was moved between the location over the soleus-Achilles tendon junction and the location over the Achilles tendon-calcaneus junction from one trial to the next. Each section of the Achilles tendon (i.e., proximal and distal) was thus imaged three times during the six isometric plantar flexion efforts.
Subjects performed six isometric plantar flexion maximum voluntary contraction (MVC) efforts slowly ramping up from 0 force to maximum force in approximately 20 s while force and ultrasound data were collected. Subjects were instructed to perform all trials in the same way. During each trial, plantar flexion forces were displayed to the subject, using an amplifier and digital display (Omega Model DP41-S). Before the start of each trial, video recording of the ultrasound images was initiated, and then a LabView data-acquisition program was invoked and synchronized with the time displayed on the monitor of the ultrasound system. Voltage outputs from the force transducer were collected at 30 Hz to match the frame rate of the video recorder. Subjects were given 2 min of rest between each trial. At the conclusion of the testing, the ultrasound probe was rotated 90° and placed over the notch of the calcaneus bone. Ultrasound images of the Achilles tendon cross-sectional area were collected at this location while the subject was at rest.
Achilles tendon force (and, therefore, triceps surae muscle force) was determined from the force-transducer data. Achilles tendon force was computed by multiplying the force determined from force-transducer data by LAR. Normalized Achilles tendon force was determined by dividing force by the MVC force obtained during the first testing session. The times during each ramping plantar flexion effort that corresponded to 0, 20, 40, 60, 80, and 100% of each subject's pretraining MVC were recorded along with the time at which new MVC efforts occurred as triceps surae muscle strength increased throughout the training program.
Achilles tendon length and strain were determined from the ultrasound images. Snappy 4.0 (by Play) was used to capture and digitize ultrasound video images that corresponded to the specified effort levels. Scion Image, an image-analysis program and PC-based version of NIH Image, was used to measure Achilles tendon lengths from the ultrasound images. Scion Image was calibrated using reference markers spaced 1 cm apart that appeared in the upper and right borders of each ultrasound image (Fig. 2). Two measurements were used to determine Achilles tendon length. D1 was defined as the distance between the center of the notch on the calcaneus bone and the center of the shadow reference (REF) created by the hollow plastic marker secured to the skin (Fig. 2). D2 was defined as the distance between REF and the soleus muscle-Achilles tendon junction (Fig. 2). Achilles tendon length for each specified effort level was calculated by summing the average D1 and D2 values. Achilles tendon rest lengths (L0), defined as the Achilles tendon length under zero muscular effort with the ankle at 90° of flexion, and length measurements occurring at the specified effort levels (Leff), were used to compute Achilles tendon strain. Achilles tendon strain at each effort level was defined as (Leff − L0)/L0, with εmax defined as the Achilles tendon strain during MVC.
Inherent in the approach used to quantify Achilles tendon length and strain are the assumptions that the subject performed each trial similarly and that the Achilles tendon moved the same relative to the skin marker reference for the probe in both the proximal and distal locations. Analysis of repeat trials yielded less than 3% coefficient of variation in the length measurements of interest, suggesting that the subjects performed each trial in a similar manner and that the length measurements were reproducible. Having a common skin marker reference within images of both the calcaneus and soleus muscle-Achilles tendon junction overcomes the measurement errors associated with other approaches cited in the literature that quantify changes in Achilles tendon length simply on the basis of the excursion of a soleus muscle-Achilles tendon junction relative to an assumed fixed calcaneus or a skin reference point (16).
Achilles tendon cross-sectional area was measured in an attempt to identify mechanisms responsible for any tendon strain changes observed. The area function in Scion Image was used to measure Achilles tendon cross-sectional area from axial ultrasound images. Determination of the accuracy of the ultrasound system for quantifying in vivo cross-sectional area measurements was not possible, but the precision of the system was determined. The precision measurements provide an indication of the magnitude of the cross-sectional area changes that could be detected during the course of the strength training program. The precision of the cross-sectional area measurement, defined as the coefficient of variation of repeat measures divided by the mean value multiplied by 100, was 10%. Achilles tendon stress was calculated by dividing Achilles tendon force by cross-sectional area. Achilles tendon stress-strain data were used to identify changes in Achilles tendon material properties.
Data were analyzed using SAS 8e for Windows (SAS Institute Inc., Cary, NC). Repeated-measures ANOVA using the general linear model was used to test for significant differences in maximum isometric triceps surae muscle strength, Achilles tendon strain, and Achilles tendon cross-sectional area between weeks (baseline, and then weeks 1, 2, 4, and 8) during the 8-wk strength training program. An alpha value of 0.017 (0.05 divided by 3 to adjust for multiple outcome tests) was used to determine statistical significance. The hypothesis was supported if there was no change or a transient change in εmax for an increased MVC triceps surae muscle force.
The group's average maximum triceps surae muscle force increased during the first 6 wk of training (23.7% above baseline at week 6) and was 21.4% above baseline at week 8 (Fig. 3). Maximum individual triceps surae muscle force changes ranged between 45.5 and −9.9%. A decrease in force was recorded in only one subject (nonnegative changes ranging between 45.5 and 3%). The absolute triceps surae muscle force (and, therefore, Achilles tendon force) developed during the maximal-effort plantar flexion efforts ranged between 1643.9 and 5019.1 N at the start of the study and between 2089.9 and 6917.4 N after 8 wk (Table 1). From baseline, a statistically significant increase in triceps surae muscle force was recorded for the group at weeks 4, 6, and 8 (P = 0.0001, P < 0.0001, and P < 0.0001, respectively).
The average percent and absolute change in maximum Achilles tendon strain was found to be statistically insignificant for the group (P = 0.607 and 0.351, respectively). Maximum Achilles tendon strain (εmax) values ranged between 0.006 and 0.135 mm·mm−1 throughout the study (Table 2).
Achilles tendon cross-sectional area ranged between 0.50 and 0.77 cm2 throughout the study. The resolution of this measurement was not sufficient to detect cross-sectional area changes during the 8-wk study; thus, an average cross-sectional area for each subject was used to compute tendon stress. The group's average maximum Achilles tendon stress was 48.6 MPa.
Average stress-strain curves were computed for the entire group (Fig. 4). Stress and strain values at effort levels corresponding to each subject's 0, 20, 40, 60, 80, and 100% MVC were combined to obtain an average stress-strain curve for the group. Average elastic modulus was calculated as the slope of the best-fit line for the stress-strain data for efforts greater than 40% of MVC. Average modulus values ranged between 1.88 and 2.89 GPa (Table 2).
The objectives of this study were 1) to further characterize the variations in tendon strains experienced in vivo, 2) to quantify the interaction between muscle strength and tendon strain through the course of a strength training program, and 3) to test the hypothesis that tendons adapt to muscle strength training to maintain strains within preferred operating ranges. Ultrasonography and joint-testing procedures were combined to achieve these objectives. The quantities used to test the hypothesis were isometric triceps surae force and Achilles tendon strain. Achilles tendon modulus and stress were also considered.
As expected, the strength training program created a significant increase in maximum isometric plantar flexion force and therefore triceps surae muscle force, which was considered equal to the force transmitted by the Achilles tendon. The group's average 8-wk increase in triceps surae muscle force, equal to 21.4%, was slightly higher than the 15.1% (24) and 19% (13) increases reported for other 8-wk training programs. The range in maximum isometric triceps surae muscle force values recorded at the end of the 8-wk program, between 2089.9 and 6917.4 N, was of similar magnitude to the range of peak Achilles tendon forces measured in a human running (between 4500 and 8500 N) (9) and humans hopping (2500 and 5500 N) (14).
The group's maximum triceps surae muscle force and Achilles tendon strain data support the hypothesis (Tables 1 and 2). There was a significant increase in maximum muscle force without an associated significant change in εmax. This suggests two things. First, the Achilles tendon seems to have a preferred strain range. If muscle force increases, then tendon properties change to limit strain. Second, on average, the Achilles tendon adapts rapidly to accommodate increased muscle force production. Temporal data pertaining to the relative adaptation of tendons to muscle strength gain is lacking. Zamora and Marini (28) overloaded the rat plantaris muscles by ablating synergist muscles and observed morphologic changes in the associated tendon after 1 and 2 wk. The tendon collagen bundles became disrupted, but there was no inflammatory response. Tendon strain during muscle contraction was not quantified, but the data suggest that the tendon would likely have strained more during a muscle contraction at week 2 compared with week 0. Specific to human strength training, Kubo et al. (13) have shown a 19% increase in muscle strength after 8 wk of strength training in young males, with no change in the deformation induced in the tendon during a maximum isometric muscle contraction. Contrasting these results, Kubo et al. (11) found the maximum strain in the vastus lateralis aponeurosis/tendon of middle-aged and elderly women changed at a statistically significant level from 11.8% (before training) to 12.5% (after training) in response to 6 months of low-resistance exercises. These differences could be attributable to differences in the age, gender, training period, and/or tendon structure of the subjects studied. Our data are consistent with those of Kubo et al. (13) and provide finer time resolution, being measured every 2 wk compared with every 8 wk. The Zamora and Marini (28) rat muscle-overload study suggested that we might see increased tendon strains within the first 2 wk, but the fact that we did not suggests that the Achilles tendon can adapt quickly to the less severe and more physiological strength training stimulus used in this study.
In addition to the absence of statistical differences in maximum Achilles tendon strain values from week to week, average weekly stress-strain curves have similar shapes. Only the stress-strain curve for week 6 is different from the others, having a shorter initial nonlinear portion causing the stress-strain curve to shift to the left (Fig. 4). Although not characterized by a statistically significant difference in maximum Achilles tendon strain, the different shape of this stress-strain curve might suggest that at 6 wk into training, there is an imbalance in the muscle tendon-adaptation response. The different "toe region" observed during week 6 could be explained by an alteration in the preload or by a shortening of the Achilles tendon. The similar slopes of the linear regions of the average stress-strain curves suggest that the material composition did not change and that, on average, damage did not likely occur in the Achilles tendon; however, it is not known whether microdamage occurred in those subjects experiencing strains that were larger than average.
The group's average maximum Achilles tendon strain did not exceed 0.056 mm·mm−1 throughout the study; however, individual Achilles tendon strain values ranged between 0.006 and 0.135 mm·mm−1 (Table 2). Although strain levels above 0.10 mm·mm−1 were previously considered to be above the rupture threshold for tendon, our findings are consistent with recent in vivo studies performed in humans and using ultrasound (10-12,17,23) that suggest that a wider range of tendon strains occur in vivo than previously thought. Average in vivo maximum strain values have been reported to range from as low as about 2.5% (17) to more than 12.5% (11). These data indicate that a wide range of tendons strains occur in vivo. The reason for such wide variations in tendon strains is not clear, but these differences may reflect differences in muscle tendon architecture and muscle-activation strategies during movement.
Two main conclusions can be drawn from this study. First, on average, the Achilles tendon seems to have a preferred strain limit that is maintained as triceps surae muscle strength increases in response to strength training. Second, Achilles tendon strain levels vary greatly depending on the individual. Reasons for these variations are unknown at this time and require further investigation.
1. Abrahams, M. Mechanical behavior of tendon in vitro: a preliminary report. Med. Biol. Eng
. 5:433-443, 1967.
2. Benedict, J., L. Walker, and E. Harris. Stress-strain characteristics and tensile strength of unembalmed human tendon. J. Biomech
. 1:53-63, 1968.
3. Butler, D., E. Grood, F. Noyes, and R. Zernicke. Biomechanics of ligaments and tendons. Exerc. Sports Sci. Rev
. 6:125-181, 1978.
4. Finni, T., J. A. Hodgson, A. M. Lai, V. R. Edgerton, and S. Sinha. Nonuniform strain of human soleus aponeurosis-tendon complex during submaximal voluntary contractions in vivo. J. Appl. Physiol
. 95:829-837, 2003.
5. Fukashiro, S., M. Itoh, Y. Ichinose, Y. Kawakami, and T. Fukunaga. Ultrasonography gives directly but noninvasively elastic characteristics of human tendon in-vivo. Eur. J. Appl. Physiol
. 71:555-557, 1995.
6. Hansen, P., P. Aagaard, M. Kjaer, B. Larsson, and S. P. Magnusson. Effect of habitual running on human Achilles tendon load-deformation properties and cross-sectional area. J. Appl. Physiol
. 95:2375-2380, 2003.
7. Huang, T. F, S. M. Perry, and L. J. Soslowsky. The effect of overuse activity on Achilles tendon in an animal model: a biomechanical study. Ann. Biomed. Eng
. 32:336-341, 2004.
8. Józsa, L., and P. Kannus. Effects of activity and inactivity in tendons. In: Human Tendons. Anatomy, Physiology and Pathology
, R. Washburn, R. Frey, N. Smith, and H. Woolsey (Eds.). Champaign, IL: Human Kinetics Publisher, pp.127-160, 1997.
9. Komi, P., S. Fukashiro, and M. Järvinen. Biomechanical loading of Achilles tendon during normal locomotion. Clin. Sports Med
. 11:521-531, 1992.
10. Kubo, K., H. Kanehisa, M. Ito, and T. Fukunaga. Effects of isometric
training on the elasticity of human tendon structures in vivo. J. Appl. Physiol
. 91:26-32, 2001.
11. Kubo, K., H. Kanehisa, M. Miyatani, M. Tachi, and T. Fukunaga. Effect of low-load resistance training on the tendon properties in middle-aged and elderly women. Acta Physiol. Scand
. 178:25-32, 2003.
12. Kubo, K., H. Kanehisa, and T. Fukunaga. Gender differences in the viscoelastic properties of tendon structures. Eur. J. Appl. Physiol
. 88:520-526, 2003.
13. Kubo, K., H. Kanehisa, and T. Fukunaga. Effects of resistance and stretching training programmes on the viscoelastic properties of human tendon structures in vivo. J. Physiol
. 538:219-226, 2002.
14. Lichtwark, G. A., and A. M. Wilson. In vivo mechanical properties of the human Achilles tendon during one-legged hopping. J. Exp. Biol
. 208:4715-4725, 2005.
15. Maganaris, C. N. Tensile properties of in vivo human tendinous tissue. J. Biomech
. 35:1019-1027, 2002.
16. Maganaris, C. N. Validity of procedures involved in ultrasound
-based measurement of human plantar flexor tendon elongation on contraction. J. Biomech
. 38:9-13, 2005.
17. Maganaris, C. N., and J. P. Paul. In vivo human tendon mechanical properties. J. Physiol
. 521:307-313, 1999.
18. Maganaris, C. N., and J. P. Paul. In vivo human tendinous tissue stretch upon maximum muscle
force generation. J. Biomech
. 33:1453-1459, 2000.
19. Maganaris, C. N., and J. P. Paul. Tensile properties of the in vivo human gastrocnemius tendon. J. Biomech
. 35:1639-1646, 2002.
20. Maganaris, C. N., M. V. Narici, and N. D. Reeves. In vivo human tendon mechanical properties: effect of resistance training in old age. J. Musculoskelet. Neuronal Interact
. 4:204-208, 2004.
21. Muramatsu, T., T. Muraoka, D. Takeshita, Y. Kawakami, Y. Hirano, and T. Fukunaga. Mechanical properties of tendon and aponeurosis of human gastrocnemius muscle
in vivo. J. Appl. Physiol
. 90:1671-1678, 2001.
22. Ochala, J., D. Lambertz, J. Van Hoecke, and M. Pousson. Effect of strength training on musculotendinous stiffness in elderly individuals. Eur. J. Appl. Physiol
. 94:126-133, 2005.
23. Reeves, N. D., C. N. Maganaris, and M. V. Narici. Effect of strength training on human patella tendon mechanical properties of older individuals. J. Physiol
. 548:971-981, 2003.
24. Rosager, S., P. Aagaard, P. Dyhre-Poulsen, K. Neergaard, M. Kjaer, and S. P. Magnusson. Load-displacement properties of the human triceps surae aponeurosis and tendon in runners and non-runners. Scand. J. Med. Sci. Sports
25. Walker, L., E. Harris, and J. Benedict. Stress-strain relationship in human cadaveric plantaris tendon: a preliminary study. Med. Electron. Biol. Eng
. 2:31-38, 1964.
26. Woo, S. L., M. A. Gomez, D. Amiel, M. A. Ritter, R. H. Gelberman, and W. H. Akeson. The effects of exercise on the biomechanical and biochemical properties of swine digital flexor tendons. J. Biomech. Eng
. 103:51-56, 1981.
27. Woo, S. L., M. Gomez, Y. K. Woo, and W. Akeson. Mechanical properties of tendons and ligaments. II. The relationship of immobilization and exercise on tissue remodeling. Biorheology
28. Zamora, A. J., and J. F. Marini. Tendon and myo-tendinous junction in an overloaded skeletal muscle
of the rat. Anat. Embryol. (Berl.)