During early swing (0%–20% of the stride cycle), the rate at which the hamstring muscles shortened increased initially and then reduced (Fig. 3, row 2). The average peak musculotendon shortening velocity at this time ranged from 0.74 ± 0.09 m·s−1 for BFSH to 1.04 ± 0.12 m·s−1 for ST (Table 2). During the middle stages of swing (20%–60% of the stride cycle), the musculotendon lengthening velocity of the hamstrings showed a characteristic biphasic pattern (Fig. 3, row). The average peak musculotendon lengthening velocity at this time ranged from −0.71 ± 0.05 m·s−1 for BFSH to −1.04 ± 0.13 m·s−1 for ST (Table 2). For BFLH, the first peak in the musculotendon lengthening velocity during swing was always larger than the second, whereas the opposite was true for BFSH. For SM and ST, the two peaks in the musculotendon lengthening velocity during swing were closer in magnitude, with some subjects having a larger first peak, whereas others displayed a larger second peak. During terminal swing (60%–75% of the stride cycle), the musculotendon shortening velocity of the hamstring muscles increased once more (Fig. 3, row 2). The average peak musculotendon shortening velocity at this time occurred just after foot-strike and ranged in magnitude from 0.16 ± 0.07 m·s−1 for BFSH to 0.69 ± 0.06 m·s−1 for ST (Table 2). The musculotendon shortening velocity of the hamstrings then decreased during the first half of stance. Whereas a musculotendon lengthening velocity occurred for BFSH during midstance for all subjects, such was not the case for the biarticular hamstring muscles. A musculotendon lengthening velocity during midstance was displayed by only three subjects for ST and by only two subjects for BFLH and SM.
The static optimization analysis predicted that SM, ST, and BFLH produced force during terminal swing and early stance, whereas BFSH produced force during terminal swing and the second half of stance (Fig. 3, row 3). Overall, there was reasonable agreement between the time during the stride cycle when the hamstring muscles were predicted to develop force and the time when the hamstring muscles were found to display EMG activity (Fig. 3, row 3 vs row 5). Average peak musculotendon force ranged from 5.49 ± 0.78 N·kg−1 for ST to 46.81 ± 6.25 N·kg−1 for SM (Table 2). In comparison to stance, peak musculotendon force produced during swing was increased by 7.2-fold, 1.5-fold, and 5.7-fold for SM, ST, and BFLH, respectively, whereas it was decreased by 0.7-fold for BFSH.
All hamstring muscles underwent a period of power absorption followed by a period of power generation during terminal swing (Fig. 3, row 4). The average peak musculotendon power absorption during terminal swing ranged from −2.60 ± 1.90 W·kg−1 for BFSH to −22.39 ± 5.40 W·kg−1 for SM, whereas the average peak musculotendon power generation during terminal swing ranged from 1.12 ± 0.69 W·kg−1 for BFSH to 7.66 ± 1.14 W·kg−1 for SM (Table 2). BFSH also underwent a period of musculotendon power absorption during terminal stance (average magnitude of −2.47 ± 2.01 W·kg−1) and a period of musculotendon power generation during early swing (average magnitude of 1.90 ± 1.06 W·kg−1), which were likely a result of the optimization algorithm preferentially distributing load onto the BFSH to control knee extension in late stance and then facilitate rapid knee flexion during early swing.
The biarticular hamstrings performed negative work only during terminal swing, whereas BFSH performed a small amount of negative work during both terminal swing and terminal stance (Table 2). All of the hamstring muscles performed more positive work during swing than during stance. The average total amount of negative work done ranged from −0.13 ± 0.05 J·kg−1 for ST to −1.06 ± 0.25 J·kg−1 for SM, whereas the average total amount of positive work done ranged from 0.10 ± 0.03 J·kg−1 for BFSH to 0.37 ± 0.07 J·kg−1 for SM. Semitendinosus was the only hamstring muscle not found to perform more negative work than positive work. In comparison to the total amount of positive work, the total amount of negative work was decreased 0.76-fold for ST and increased by 2.86-fold for SM.
The purposes of the present study were to quantify the biomechanical load (i.e., musculotendon strain, velocity, force, power, and work) experienced by the hamstrings across a full stride cycle during overground sprinting and to compare the biomechanical load across the four hamstring muscles (i.e., SM, ST, BFLH, and BFSH). The main outcomes can be summarized as follows. First, the biarticular hamstrings undergo a stretch–shortening cycle during sprinting, with the lengthening phase occurring during terminal swing and shortening phase commencing just before foot-strike and continuing throughout stance. The timing of this stretch–shortening cycle corresponds with the period of hamstring EMG activity as measured in this study (Fig. 3, row 5) and in many previous studies (7,21,23,25,26,35,45). Second, the biomechanical load on the biarticular hamstring muscles was found to be greatest during terminal swing. At this time in the stride cycle, ST, SM, and BFLH all reached peak musculotendon strain, produced peak musculotendon force, and performed much negative work. Third, when comparing the various hamstring muscles, the following results were found: (i) BFLH had the largest peak musculotendon strain (12.0% increase in length from upright stance position), (ii) ST displayed the greatest musculotendon lengthening velocity, and (iii) SM produced the highest musculotendon force, absorbed and generated the most musculotendon power, and performed the largest amount of positive and negative work. These outcomes have implications for improving existing understanding of the pathomechanics of sprinting-related hamstring muscle strain-type injuries.
When comparing data from the present study with that from previous studies, the majority of the observed discrepancies are most likely attributable to the combined effect of three main factors. First, differences in the procedures used to record experimental data. In particular, several prior studies have measured data while sprinting on a treadmill (6,7,39), whereas data from the present study as well as those from Schache et al. (33) were measured while sprinting overground. Frishberg (14) compared treadmill and overground sprinting at 9.2 m·s−1 for five male athletes and found several trunk and lower limb kinematic parameters to be different between the two conditions. Second, differences in the computational approach adopted to estimate muscle forces. Thelen et al. (39) and Chumanov et al. (6,7) used a forward dynamics simulation approach (38), whereas Schache et al. (33) and the present study used an inverse dynamics–based static optimization approach (3). All studies have applied the same general performance criterion (i.e., minimization of the sum of the square of muscle activations ). However, whereas most studies have not imposed additional constraints on when, in the stride cycle, muscle activations can be predicted to occur, Chumanov et al. (7) explicitly constrained the hamstring and rectus femoris muscles to be inactive for a brief period during terminal stance and early swing. Third, variability in the maximum sprinting speeds investigated may explain, to some degree, discrepancies in reported results. The average sprinting speed in the present study was 9.0 m·s−1, whereas Schache et al. (33) recorded data at 7.4 m·s−1 and Chumanov et al. (7) used average maximum speeds of 8.0 and 7.1 m·s−1 for males and females, respectively. It is therefore possible that the larger hamstrings muscle forces in the present study compared with those in Schache et al. (33) and Chumanov et al. (7) are at least partially attributable to differences in test speeds. However, other studies (e.g., Thelen et al.  and Chumanov et al. ) involved maximum sprinting speeds that were similar in magnitude to the present study; hence, contrasting results in this instance cannot be explained on the basis of test speeds alone.
Despite the methodological differences among studies conducted to date, there is remarkable qualitative consistency in the reported findings. All studies have found the biarticular hamstrings to be subjected to large loads during the terminal swing phase of sprinting (6,7,33,39), with such loads exceeding those during stance (7,33). Furthermore, when comparing the various hamstrings, it has been found that BFLH experiences the greatest musculotendon strain with respect to upright stance (6,7,34,40) and develops its peak musculotendon force slightly earlier in swing (6), whereas SM generates the largest peak musculotendon force and performs the greatest amount of work (6,7). These findings are all in agreement with the main outcomes from the present study. A quantitative comparison of results obtained from the various studies is given in Table 3. In some instances, data are reasonably similar; for example, the negative work done by SM during sprinting. However, for other parameters such as peak stance musculotendon force, data from the present study differ from equivalent data presented by Chumanov et al. (7) (see below).
Musculoskeletal modeling is the only practicable method for determining quantities such as muscle force, power, and work done under in vivo conditions (29). There is much evidence in the literature to show that computer-based models of the musculoskeletal system accurately predict biomechanical behavior. For instance, studies have compared model-derived estimates of hip- and knee-joint contact forces against simultaneously recorded in vivo data from instrumented (strain-gauged) joint implants for common activities of daily living, including walking (19,24) and stair climbing (19). These studies presented calculated joint contact forces that can be considered to be in good agreement with those directly measured from the instrumented joint implants. Although generating accurate estimates of joint contact forces is a necessary, but not sufficient, condition for concluding that the corresponding lower limb muscle forces are also accurate, confidence in model-derived estimates can be further increased by qualitatively comparing patterns of muscle forces with measured EMG data. In this regard, several studies have demonstrated that an inverse dynamics–based static optimization approach, as was used in the present study, is capable of producing lower limb muscle forces that have similar time histories to experimental measurements of EMG data for walking at the preferred speed (3) and running (16,29).
Unfortunately, as quantities such as muscle force, power, and work done cannot be measured in vivo by noninvasive means, it is not possible to directly validate estimates of hamstrings biomechanical load obtained in the present study. However, there are numerous factors that provide indirect evidence to indicate that the results from this study are reasonable. First, inverse dynamics–based joint moments in the present study are qualitatively and quantitatively consistent with equivalent data reported in the literature (1,32). Second, the estimation of large biomechanical loads on the biarticular hamstring muscles during the terminal swing phase of sprinting is in agreement with EMG studies that have reported peak activity of the medial and lateral hamstrings to occur during terminal swing (21,26,45). Third, sagittal plane moment arms for each of the hamstring muscles in the model plotted against hip- and knee-joint angle are consistent with experimental data from studies that have directly measured hamstrings moment arms using cadaveric specimens (see Figures, Supplemental Digital Contents 1 and 2, Sagittal plane moment arms at the hip and knee joint for each of the hamstring muscles in the model compared with available experimental data from studies that have directly measured hamstrings moment arms using cadaveric specimens; https://links.lww.com/MSS/A122, https://links.lww.com/MSS/A123). Finally, the estimated distribution of musculotendon force across the various hamstring muscles is consistent with their force-generating capacity based on muscle architecture. The distribution of musculotendon force across the hamstring muscles was primarily determined by the maximum isometric force (F0M) attributed to each muscle in the model. A given muscle’s F0M is proportional to its physiological cross-sectional area (PCSA), which is a measure of the number of parallel muscle fibers acting within a muscle. Muscles with a larger PCSA have a greater force-generating capacity (41,42). Studies have used cadavers to directly calculate PCSA for the various hamstring muscles (41,42). Ward et al. (41) and Wickiewicz et al. (42) reported PCSA to range from 16.9 to 18.4 cm2 and 4.8 to 5.4 cm2 for SM and ST, respectively. Ward et al. (41) measured a PCSA of 11.3 cm2 and 5.1 cm2 for BFLH and BFSH, respectively, whereas Wickiewicz et al. (42) measured a PCSA of 12.8 cm2 for the entire biceps femoris muscle. Taken together, these data indicate that the force-generating capacity of the hamstring muscles is greatest for SM and BFLH and smallest for ST and BFSH. Furthermore, the force-generating capacity of SM exceeds that of BFLH. The relative distribution of musculotendon force across the various hamstring muscles calculated in the present study is therefore consistent with what would be expected on the basis of reported PCSA measurements.
The present study was associated with several limitations and assumptions. First, only a single representative trial was analyzed per subject. It is acknowledged that it would have been ideal to have analyzed multiple trials for each subject. However, for a trial to be deemed successful, subjects were required to complete a full stride cycle of valid force plate contacts for the designated lower limb of interest. It typically required several attempts to obtain a representative trial. To avoid the potential confounding effect of fatigue, the study was therefore limited to a single trial per subject. Second, the outputs of the musculoskeletal model are sensitive to the values assumed for the musculotendon force–length–velocity properties. Such properties were not directly measured for each participant in the present study; rather literature-derived estimates were used (Table 1). However, every effort was made to obtain data that were considered reasonable. For example, pennation angle (α0) for each individual hamstring muscle matched mean data reported by Ward et al. (41). Furthermore, optimal muscle fiber lengths (l0M) for SM, ST, and BFLH were based on data reported by Wickiewicz et al. (42), whereas l0M for BFSH was based on data reported by Ward et al. (41). Where it was not possible to source direct measurements of certain hamstrings properties, values were based on recommendations reported by others: data for tendon compliance ([Latin Small Letter Open E]0T) were obtained from Zajac (46), whereas data for tendon slack length (lsT) were obtained from Delp et al. (11). It is also worth noting in this regard that the magnitudes of certain musculotendon force–length–velocity properties are far more critical when the relative dynamics within the musculotendon unit is of key interest, i.e., quantifying the dynamics of the tendon versus muscle components. Thelen et al. (39) have previously demonstrated that incremental variations in the value assumed for tendon compliance ([Latin Small Letter Open E]0T in Table 1) substantially influences predictions of BFLH tendon stretch and work done during sprinting. Analyses in the present study were therefore restricted to the quantification of net musculotendon dynamics, which is unfortunate because it is likely that tendon function in vivo has an important role in fast movements such as sprinting by storing and releasing energy.
Third, estimates of hamstrings musculotendon force, power, and work done in this study are also limited to the particular method for calculating these parameters. Musculotendon forces were calculated using inverse dynamics–based static optimization. This approach is computationally efficient and has been commonly applied to estimate lower limb muscle forces during locomotion (3,9,16,24,29). However, unlike a dynamic optimization algorithm, static optimization neglects muscle activation dynamics (3). Although it has been shown that static and dynamic optimization algorithms yield similar results when applied to walking (3) and slower speeds of running (29), it is not currently known whether these different computational methods generate consistent results for faster locomotion speeds (i.e., sprinting). The performance criterion used to solve the optimization problem was the sum of the square of muscle activations (3). Although this specific criterion has been applied by previous researchers to compute lower limb muscle forces during sprinting (6,7,33,39), it is acknowledged that minimization of muscle stress during sprinting may not be the most important consideration. Furthermore, such performance criteria may be limited in their capacity to predict cocontraction among antagonistic pairs of muscles (8).
One parameter that would seem to have been somewhat underestimated in this study was peak stance musculotendon force for the biarticular hamstrings. Only BFSH was predicted to be producing force in the second half of stance. As previously mentioned, the magnitude of peak stance hamstrings musculotendon force from the current study is less than that from Chumanov et al. (7). Peak stance hamstrings force ranged from 3.64 N·kg−1 for ST to 14.78 N·kg−1 for BFSH in this study, whereas it ranged from 6.2 N·kg−1 for ST to 12.1 N·kg−1 for SM in the study by Chumanov et al. (7) (Table 3). Prior investigations recording lower limb muscle EMG activity during sprinting have found the medial and lateral hamstrings to be active throughout stance (7,21,23,25,26,35,45). Similarly, EMG activity measured from the medial and lateral hamstrings in this study was found to commence during terminal swing and continue throughout stance until the start of initial swing (Fig. 3, row 5). Although the relationship between EMG and muscle force for fast dynamic contractions is complicated and affected by many factors (12), such experimental data would suggest that, for the biarticular hamstrings during stance, the predicted magnitude of musculotendon force in the present study may be lower than what would be expected.
This inconsistency is most likely attributable to the computational approach used to calculate muscle forces; that is, the inability of inverse dynamics–based static optimization when combined with a minimum-stress performance criterion to adequately predict antagonistic cocontraction. Evidence for this assertion is provided by Collins (8), who evaluated the performance of a variety of optimization algorithms in calculating muscle forces during walking. Although minimization of the sum of muscle activations squared was not specifically tested, optimization algorithms that were included were all found to be particularly insensitive to the prediction of antagonistic quadriceps–hamstrings activity during stance. Predicting high levels of biarticular hamstrings activity throughout stance when there is a net hip flexor moment during the second half of stance and a net knee extensor moment for the majority of stance would not be (from a computational perspective) the most energy-efficient way to distribute the joint moments across the various lower limb muscles. Large muscle forces from the biarticular hamstrings during the first half of stance would likely require greater activations to be computed from the knee extensor muscles, so as to counter the unwanted mechanical effect of the hamstrings at the knee joint and maintain equality with the inverse dynamics–based joint moments. Hence, the hip extensor moment during the first half of stance was preferentially distributed onto the gluteus maximus muscle. In a similar manner, large muscle forces from the biarticular hamstrings during the second half of stance would likely require greater activations to be computed from the hip flexor and knee extensor muscles to maintain equality with the inverse dynamics moments.
Cocontraction of antagonistic muscles can be used to modulate the impedance and thus stability of a joint, which would seem advantageous during the stance phase of sprinting when the lower limb is subjected to a high-frequency impact force. Rather interestingly, it has been demonstrated that people who have learned to cope with compromised knee joint stability (i.e., anterior cruciate ligament deficiency) display increased stance phase hamstrings EMG activity during locomotion in comparison to a group of healthy counterparts (5). It is therefore speculated that after foot-strike during sprinting, the hamstrings are active not to counter the external moments generated largely by the GRF, but rather to provide alternative functions, such as joint stability and/or proprioception.
Because the biarticular hamstring muscles all reached peak musculotendon strain, produced peak musculotendon force, and performed much negative work during the terminal swing phase of sprinting, it would seem that the hamstrings are likely to be most vulnerable to injury at this time in the stride cycle. Unlike concentric contractions, eccentric contractions have been shown to be capable of producing muscle fiber damage (28). Furthermore, the conclusion that the hamstrings are likely to be at greater risk of injury during terminal swing as opposed to the stance phase concurs with the findings from two recently published, yet independent, case reports that unexpectedly captured biomechanical data of a running athlete suffering a hamstring muscle strain-type injury (18,34). Both of these studies identified terminal swing as the period in the stride cycle when the injury stimulus most likely occurred.
Most hamstring muscle strain injuries involve the biceps femoris muscle (4). It was therefore of interest to determine whether there are biomechanical reasons for this clinical observation; hence parameters such as musculotendon strain, velocity, force, power, and work were compared for each individual hamstring muscle. Based on data from the present study as well as data from previous studies (6,7), the propensity for hamstring muscle strain-type injuries to frequently involve biceps femoris cannot be simply explained on the basis of peak force or the total amount of negative work done because both of these parameters were estimated to be greatest for SM not BFLH or BFSH. In accordance with Thelen et al. (40) and Chumanov et al. (6,7), the amount of musculotendon strain (lengthening with respect to upright stance) was found to distinguish BFLH from SM and ST. An average peak strain of 12.0% was experienced by BFLH during sprinting, which was 2.2% and 3.3% greater than that for SM and ST, respectively (Table 2). These data therefore indicate that the degree of musculotendon strain may be the more relevant parameter in understanding the apparent vulnerability of biceps femoris to injury. This conclusion is consistent with the results from an animal-based study, whereby muscle damage after an eccentric contraction was found not simply to be a function of peak muscle force but rather was due to the magnitude of the strain experienced by the musculotendon unit during contraction (28).
In summary, the present study found the biarticular hamstrings (SM, ST, and BFLH) to be lengthening, producing peak force, and performing much negative work (energy absorption) during the terminal swing phase of the stride cycle. This study also found differing biomechanical loads for each individual hamstring muscle: BFLH exhibited the largest peak strain, ST displayed the greatest lengthening velocity, and SM produced the highest peak force, absorbed and generated the most power, and performed the largest amount of positive and negative work. As peak musculotendon force and strain for BFLH, ST, and SM occurred around the same time during terminal swing, it is suggested that this period in the stride cycle may be when the biarticular hamstrings are at greatest injury risk. On this basis, it is recommended that hamstring injury prevention or rehabilitation programs should be preferentially biased toward strengthening exercises that primarily involve eccentric contractions performed with high loads at longer musculotendon lengths.
Financial support for this project was provided by the Physiotherapy Research Foundation Tagged Sports Physiotherapy Australia research grant (T08-THE/SPA(1)018), the Australian Research Council Discovery Projects grant DP0772838, and the Australian Research Council Linkage Projects grant LP110100262.
The authors thank Doug Rosemond for his assistance with experimental data collection.
The authors have no conflicts of interest to declare.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Ae M, Miyashita K, Yokoi T, Hashihara Y. Mechanical power and work done by the muscles of the lower limb during running at different speeds. In: Jonsson B, editor. Biomechanics X-B. Champaign (IL): Human Kinetics Publishers, Inc; 1987. p. 895–9.
2. Anderson FC, Pandy MG. Dynamic optimization of human walking. J Biomech Eng. 2001; 123: 381–90.
3. Anderson FC, Pandy MG. Static and dynamic optimization solutions for gait are practically equivalent. J Biomech. 2001; 34: 153–61.
4. Askling CM, Tengvar M, Saartok T, Thorstensson A. Acute first-time hamstring strains during high-speed running. Am J Sports Med. 2007; 35: 197–206.
5. Boerboom AL, Hof AL, Halbertsma JPK, et al.. Atypical hamstrings electromyographic activity as a compensatory mechanism in anterior cruciate ligament deficiency. Knee Surg Sports Traumatol Arthrosc. 2001; 9: 211–6.
6. Chumanov ES, Heiderscheit BC, Thelen DG. The effect of speed and influence of individual muscles on hamstring mechanics during the swing phase of sprinting. J Biomech. 2007; 40: 3555–62.
7. Chumanov ES, Heiderscheit BC, Thelen DG. Hamstring musculotendon dynamics during stance and swing phases of high-speed running. Med Sci Sports Exerc. 2011; 43 (3): 525–32.
8. Collins JJ. The redundant nature of locomotor optimization laws. J Biomech. 1995; 28: 251–67.
9. Crowninshield RD, Brand RA. A physiologically based criterion of muscle force prediction in locomotion. J Biomech. 1981; 14: 793–801.
10. Delp SL, Anderson FC, Arnold AS, et al.. OpenSim: open-source software to create and analyze dynamic simulations of movement. IEEE Trans Biomed Eng. 2007; 54: 1940–50.
11. Delp SL, Loan P, Hoy MG, Zajac FE, Topp EL, Rosen JM. An interactive graphics-based model of the lower extremity to study orthopaedic surgical procedures. IEEE Trans Biomed Eng. 1990; 37: 757–67.
12. Disselhorst-Klug C, Schmitz-Rode T, Rau G. Surface electromyo graphy and muscle force: limits in sEMG–force relationship and new approaches for applications. Clin Biomech. 2009; 24: 225–35.
13. Ekstrand J, Hagglund M, Walden M. Epidemiology of muscle injuries in professional football (soccer). Am J Sports Med. 2011; 39: 1226–32.
14. Frishberg BA. An analysis of overground and treadmill sprinting. Med Sci Sports Exerc. 1983; 15 (6): 478–85.
15. Fuller CW, Laborde F, Leather RJ, Molloy MG. International rugby board rugby world cup 2007 injury surveillance study. Br J Sports Med. 2008; 42: 452–9.
16. Glitsch U, Baumann W. The three-dimensional determination of internal loads in the lower extremity. J Biomech. 1997; 30: 1123–11.
17. Hamner SR, Seth A, Delp SL. Muscle contributions to propulsion and support during running. J Biomech. 2010; 43: 2709–16.
18. Heiderscheit BC, Hoerth DM, Chumanov ES, Swanson SC, Thelen BJ, Thelen DG. Identifying the time of occurrence of a hamstring strain injury during treadmill running: a case study. Clin Biomech. 2005; 20: 1072–8.
19. Heller MO, Bergmann G, Deuretzbacher G, et al.. Musculo-skeletal loading conditions at the hip during walking and stair climbing. J Biomech. 2001; 34: 883–93.
20. Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol. 2000; 10: 361–71.
21. Higashihara A, Ono T, Kubota J, Okuwaki T, Fukubayashi T. Functional differences in the activity of the hamstring muscles with increasing running speed. J Sports Sci. 2010; 28: 1085–92.
22. Holzbaur KR, Murray WM, Delp SL. A model of the upper extremity for simulating musculoskeletal surgery and analyzing neuromuscular control. Ann Biomed Eng. 2005; 33: 829–40.
23. Jonhagen S, Ericson MO, Nemeth G, Eriksson E. Amplitude and timing of electromyographic activity during sprinting. Scand J Med Sci Sports. 1996; 6: 15–21.
24. Kim H-J, Fernandez JW, Akbarshahi M, Walter JP, Fregly BJ, Pandy MG. Evaluation of predicted knee joint muscle forces during gait using an instrumented knee implant. J Orthop Res. 2009; 27: 1326–11.
25. Kyröläinen H, Avela J, Komi PV. Changes in muscle activity with increasing running speed. J Sports Sci. 2005; 23: 1101–9.
26. Kyröläinen H, Komi PV, Belli A. Changes in muscle activity patterns and kinetics with increasing running speed. J Strength Cond Res. 1999; 13: 400–6.
27. Li X, Zhou P, Aruin AS. Taeger–Kaiser energy operation of surface EMG improves muscle activity onset detection. Ann Biomed Eng. 2007; 35.
28. Lieber RL, Friden J. Muscle damage is not a function of muscle force but active muscle strain. J Appl Physiol. 1993; 74: 520–6.
29. Pandy MG, Andriacchi TP. Muscle and joint function in human locomotion. Annu Rev Biomed Eng. 2010; 12: 401–33.
30. Pedersen DR, Brand RA, Cheng C, Arora JS. Direct comparison of muscle force predictions using linear and non-linear programming. J Biomech Eng. 1987; 109: 192–9.
31. Reinschmidt C, Van Den Bogert AJ, Murphy N, Lundberg A, Nigg BM. Tibiocalcaneal motion during running, measured with external and bone markers. Clin Biomech. 1997; 12: 8–16.
32. Schache AG, Blanch PD, Dorn TW, Brown NAT, Rosemond D, Pandy MG. Effect of running speed on lower limb joint kinetics. Med Sci Sports Exerc. 2011; 43 (7): 525–32.
33. Schache AG, Kim H-J, Morgan DL, Pandy MG. Hamstring muscle forces prior to and immediately following an acute sprinting-related muscle strain injury. Gait Posture. 2010; 32: 136–40.
34. Schache AG, Wrigley TV, Baker R, Pandy MG. Biomechanical response to hamstring muscle strain injury. Gait Posture. 2009; 29: 332–8.
35. Simonsen EB, Thomsen L, Klausen K. Activity of mono- and biarticular leg muscles during sprint running. Eur J Appl Physiol Occup Physiol. 1985; 54: 524–32.
36. Solnik S, Rider P, Steinweg K, Devita P, Hortobagyi T. Taeger–Kaiser energy operator signal conditioning improves EMG onset detection. Eur J Appl Physiol. 2010; 110: 489–98.
37. Stacoff A, Reinschmidt C, Stüssi E. The movement of the heel within a running shoe. Med Sci Sports Exerc. 1992; 24 (6): 695–701.
38. Thelen DG, Anderson FC. Using computed muscle control to generate forward dynamic simulations of human walking from experimental data. J Biomech. 2006; 39: 1107–15.
39. Thelen DG, Chumanov ES, Best TM, Swanson SC, Heiderscheit BC. Simulation of biceps femoris musculotendon mechanics during the swing phase of sprinting. Med Sci Sports Exerc. 2005; 37 (11): 1911–8.
40. Thelen DG, Chumanov ES, Hoerth DM, et al.. Hamstring muscle kinematics during treadmill sprinting. Med Sci Sports Exerc. 2005; 37 (1): 108–14.
41. Ward SR, Eng CM, Smallwood LH, Lieber RL. Are current measurements of lower extremity muscle architecture accurate? Clin Orthop Rel Res. 2009; 467: 1074–82.
42. Wickiewicz TL, Roy RR, Powell PL, Edgerton VR. Muscle architecture of the human lower limb. Clin Orthop Rel Res. 1983; 179: 275–83.
43. Wood GA. Biomechanical limitations to sprint running. In: Van Gheluwe B, Atha J, editors. Current Research in Sports Biomechanics. Basel (Switzerland): Karger; 1987. p. 58–71.
44. Yamaguchi GT, Zajac FE. A planar model of the knee joint to characterise the knee extensor mechanism. J Biomech. 1989; 22: 1–10.
45. Yu B, Queen RM, Abbey AN, Liu Y, Moorman CT, Garrett WE. Hamstring muscle kinematics and activation during overground sprinting. J Biomech. 2008; 41: 3121–6.
46. Zajac FE. Muscle and tendon: properties, models, scaling, and application to biomechanics and motor control. In: Bourne JR, editor. CRC Critical Reviews in Biomedical Engineering. Baca Raton (FL): CRC Press; 1989. p. 359–411.
47. Zajac FE. Muscle coordination of movement: a perspective. J Biomech. 1993; 26: 109–24.
Keywords:©2012The American College of Sports Medicine
RUNNING BIOMECHANICS; MUSCLE FORCE; MUSCLE FUNCTION; MUSCLE INJURY; MUSCULOSKELETAL MODELING