Journal Logo

Applied Sciences: Biodynamics

The role of mechanical and neural restraints to joint range of motion during passive stretch


Author Information
Medicine & Science in Sports & Exercise: June 1998 - Volume 30 - Issue 6 - p 928-932
  • Free


Musculoskeletal flexibility is typically characterized by maximum range of motion (ROM) in a joint or series of joints (19). Determinants of maximum ROM and, therefore, determinants of flexibility are not well understood. A neural component has been proposed whereby passive stretch elicits a reflex response facilitating greater active contractile resistance to stretch at the extremes of motion (5,16). A reflex response to slow passive stretch has been demonstrated in several studies (22-25), but the relative contribution to resistance to stretch was not determined. Clinical research on humans has emphasized the role of neural adaptations in both acute and chronic improvements in ROM (6,10,13,26). However, the neural theory of flexibility has not been substantiated (23-25) and remains controversial (6,15).

In contrast, a viscoelastic component of flexibility has been proposed whereby stretch is limited by passive mechanical forces of muscle tissue (elastic response), and acute adaptations to sustained or repeated stretch (creep and stress relaxation) are a function of the viscous response (27). Viscoelastic behavior of human skeletal muscle has been demonstrated in vivo in healthy individuals (20,22,30) and in patients with neurological disorders (14,29). Passive mechanical forces have been quantified during passive stretch of skeletal muscle according to the relationship of change in ROM to change in torque, in the absence of EMG activity (7,8,11,12,14,29,30). Although resistance to stretch in the mid-range of motion seems to be a function of passive mechanical forces (18,27), an active contractile response may contribute resistance at terminal ROM.

Most clinical tests of musculoskeletal flexibility are based on a determination of maximum ROM. For example, the straight leg raise test (SLR) measures passive hip flexion ROM with the knee in extension and is a standard goniometric test of hamstring flexibility (19). It is not known if the SLR test measures passive mechanical restraints to joint motion or contractile resistance to stretch at terminal ROM. It is possible that individuals with a lower maximum ROM have a greater stretch-induced contractile response to stretch preventing them from reaching the anatomic end of ROM. However, if passive mechanical restraints limit SLR, ROM individuals with lower maximum ROM should exhibit greater resistance to stretch throughout the ROM. Therefore, the increase in torque over a fixed ROM, in the absence of EMG, should be greater in individuals with a lower maximum ROM. The purpose of this study was to examine whether maximum SLR ROM was limited by passive mechanical forces or stretch-induced neural responses to stretch.


Test Protocol

Sixteen subjects (8 men, 8 women; age 29.1 ± 1.5 yr) volunteered for participation in this study. The subjects were recreational athletes without orthopedic injury. The protocol involved initially recording the EMG activity for maximal voluntary contractions (MVC) of the quadriceps and hamstring muscle groups. An instrumented SLR test was then applied to the right limb of each subject. Hip flexion ROM, quadriceps, and hamstring EMG activity and resistance to stretch were continuously measured. Finally, a standard goniometric measure of maximum SLR ROM was made by a physical therapist (PT).


Instrumented SLR. A hydraulically powered motorized rotating frame was developed (Scientific Stretching, Nova Scotia, Canada) for applying the instrumented SLR stretch. The axis of rotation was vertically adjustable to align with the subject's hip joint. The limb was attached at the ankle, by a chain in series with a load cell, to a crossbar that was adjustable to the subject's limb length. The instrumented SLR stretch was applied to the right limb, ending when the subject reported discomfort. The knee was braced with a postsurgical knee immobilizer locked in extension. During the stretch, the contralateral limb was fixed to the table with a strap to limit pelvic rotation.

Range of motion measurement. An electrogoniometer (Penny & Giles, Gwent, United Kingdom) was placed over the hip joint. The distal endblock was aligned with the femur and attached directly to the skin. The proximal endblock was attached to a brick and aligned with the subject's trunk. The electrogoniometer had a sensitivity of 1°. ROM is reported in degrees of hip flexion with the knee braced in extension. Following the instrumented SLR test, the limb was brought again to the same final ROM (as measured by the electrogoniometer), and hip flexion ROM was measured manually with a standard goniometer for comparison.

Measurement of resistance to stretch. The force required to raise the lower extremity was measured by a load cell (Kistler Instruments, Winterthur, Switzerland) in series with a chain attached to the subject's ankle. Direction of pull was kept at 90 degrees to the leg throughout the ROM by adjusting the length of the moving arm for the subject's limb length. The load cell was calibrated before testing with a 113.27 N load (25 lb). During testing the range in peak forces measured during the passive stretches was 38 to 110 N. The tension measurement had a sensitivity of 0.4 N.

In this model, the measured resistance to stretch is a gross force representing the weight of the lower extremity at the given hip flexion angle (limb weight) plus the resistance to stretch (net force). The net force represents the tensile force applied to the hamstring muscle group (resistance to stretch) and was estimated as follows: Net Force = Gross Force − [cos (hip flexion ROM) × Limb Weight].

Limb weight was defined as the force required to hold the lower extremity horizontally with the hip in neutral position. The limb was held at this point for approximately 3 s to set the electrogoniometer at zero. The net force measurements were converted to torque based on the limb length (greater trochanter to lateral malleolus) and are reported as Newtonmeters (N·m). This technique for limb weight correction during SLR measurements has been used previously (11,12,22).

EMG measurement. A pair of Ag/AgCl surface electrodes 1 cm in diameter were placed approximately 6 cm apart midway between the gluteal fold and the knee joint in the line of the long head of the biceps femoris muscle. These electrode placements were chosen to provide a gross indication of stretch-induced hamstring activity without distinguishing between the activity of individual hamstring synergists. A second pair of electrodes was placed midway between the anterior superior iliac spine (ASIS) and the superior border of the patella in the line of the rectus femoris muscle. These electrode placements were chosen because the rectus femoris is the prime mover during the SLR test, and it was necessary to ensure that the subject did not actively assist the maneuver. The skin was shaved, cleaned, and abraided at each site before attaching electrodes.

The raw EMG signal was recorded by telemetry (Noraxon, Scottsdale, AZ) with a bandwidth of 0-500 Hz and a CMRR of 135 dB. Before administering the instrumented SLR stretch, the EMG signal from the quadriceps and hamstring muscle groups was recorded during maximum voluntary isometric contractions (MVC) of the hip flexors and hip extensors in the SLR test position. The limb was held by the tester in approximately 10° of hip flexion, and the subject was asked to maximally contract against the tester's fixed resistance in the direction of hip flexion and then in the direction of hip extension. The knee was braced for MVC testing, as it was during the stretch procedure. Each contraction lasted for approximately 3 s. The peak activity over 1 s was full-wave rectified and integrated (μV·s). During the instrumented SLR stretch, the raw peak hamstring EMG activity was full-wave rectified and integrated for a 1-s interval. The stretch-induced hamstring IEMG was quantified as a percent of hamstring MVC. Rectus femoris activity was quantified similarly.

Instrumented SLR data processing. The force and ROM signals were digitized at 1000 Hz and upon analysis were resampled to 100 Hz to facilitate calculations. Signals were low pass filtered with a cutoff of 20 Hz using a 40th order FIR filter. Noise spikes in the data were removed using a median filter. Force units were converted from mV to N based on the pretest calibration and then converted to torque (N·m) based on the subject's limb length. Over the period of data collection, the calibration range was 0.423 N·mV−1 to 0.376 N·mV−1.

During the limb weight measurement, the electrogoniometer was set at zero. During the SLR test, maximum ROM was noted visually from the digital display of the electrogoniometer. ROM units were converted from voltage to degrees based on the voltage differential from zero to maximum ROM. The voltage outputs for the load cell and electrogoniometer were linear for the ranges used.

Torque/ROM curves were calculated for zero ROM to maximum ROM and back to zero ROM (The MathWorks Inc., Natick, MA). The area under the curve from zero to maximum ROM was calculated and is referred to as total energy absorbed. The difference between the area from zero to maximum ROM and the area from maximum ROM back to zero was computed and is referred to as energy loss during the stretch. Torque at maximum ROM is referred to as peak torque. Additionally, torque/ROM changes were analyzed over a common ROM to facilitate interindividual comparisons. The increase in torque and the area under the curve from (energy absorbed) 20-50° were quantified for each subject.

Clinical SLR. The clinical measure of maximum SLR ROM was made with a standard goniometer by a physical therapist. The PT was blinded to the ROM achieved during the instrumented test. The limb was manually raised while held in extension. The contralateral leg was manually restrained to limit pelvic rotation. The goniometer axis was placed on the greater trochanter with the stable arm aligned with the trunk parallel to the table, and the mobile arm aligned with the femur toward the lateral femoral condyle. Maximum ROM was determined subjectively by the tester using a combination of detection of movement from the contralateral ASIS and a muscular end feel. This measurement was always made after the instrumented SLR measurement either on the same day or the following day.

Statistical analysis. The clinical SLR measurement (PT SLR ROM) was used as the criterion measurement of hamstring flexibility. Linear regression was used to determine the relationship between PT SLR ROM and (1) total energy absorbed, (2) energy loss, (3) peak torque, (4) increase in torque from 20 to 50°, and (5) energy absorbed from 20 to 50° and stretch-induced IEMG. Stepwise multiple regression analysis was used to determine if a combination of variables improved the association with PT SLR ROM. Descriptive results are reported as mean ± standard error of the mean.


Torque/ROM curves (Fig. 1) showed a typical exponential relationship with significant energy loss in all subjects. Total energy absorbed was 26.4 ± 2.7 N·m·rad−1; energy loss was 57 ± 3%; peak torque was 55.8 ± 4.9 N·m, increase in torque from 20 to 50° was 13.2 ± 1.8 N·m, and the energy absorbed from 20 to 50° was 6.7 ± 0.7 N·m·rad−1. PT SLR ROM was positively related to total energy absorbed (r = 0.49, P = 0.044), negatively related to the increase in torque from 20 to 50° (r = −0.81, P < 0.0001; Fig. 2), and negatively related to energy absorbed from 20 to 50° (r = −0.73, P < 0.001). PT SLR ROM was unrelated to energy loss (r = 0.09, P = 0.75), peak torque (r = 0.09, P = 0.72), and stretch-induced hamstring activity (r = −0.06, P = 0.8). With stepwise multiple regression analysis, a combination of the increase in torque from 20 to 50° (negatively related) and total energy absorbed (positively related) improved the relationship to PT SLR ROM (r = 0.89, P = 0.001).

Figure 1
Figure 1:
Typical torque/ROM curve showing an exponential relationship with significant energy loss. Total energy absorbed (area under curve from "Start" to "Max ROM") = 28.6 N·m·rad−1. Energy loss (area under curve from "Max ROM" to "End"/total energy absorbed) = 47%. Increase in torque from 20 to 50° = 17.4 N·m. Energy absorbed from 20 to 50° = 8.0 N·m·rad−1. Arrows indicate the phases corresponding with raising and lowering the limb.
Figure 2
Figure 2:
PT SLR ROM versus increase in torque from 20 to 50°; y = 38.3-0.26x; r2 = 0.66; P < 0.0001 (dotted curve = 95% confidence intervals).

During the instrumented SLR test, there was also good agreement between maximum ROM measured by the electrogoniometer and that measured by the standard goniometric measurement (83 ± 5° and 84 ± 5°; r = 0.97, P < 0.001). PT SLR ROM (98 ± 5°) was significantly greater than the instrumented SLR ROM (83 ± 5°; P < 0.001), but the two ROM measurements were highly correlated (r = 0.93, P < 0.001). During instrumented SLR, testing the rate of stretch was 3.1 ± 0.2°·s−1 for raising the limb and 1.7 ± 1°·s−1 for lowering the limb (P = 0.001). This difference was unintentional and a function of the hydraulic system used.

All discernible EMG activity in both muscle groups occurred at maximum ROM with electrical silence throughout the rest of the ROM in all subjects. IEMG activity was quantified over the 1 s before maximum ROM. Minimal stretch-induced EMG activity was elicited from the hamstring muscle group (3 ± 1% MVC) with a similar level of EMG activity elicited from the quadriceps muscle groups (4 ± 2% MVC). Eight subjects had no discernible hamstring activity, with 6 ± 2% MVC in the other subjects. PT SLR ROM was 99 ± 9° in patients with no discernible hamstring activity and 96 ± 8° in patients who elicited a hamstring response to stretch (P = 0.8). Torque/ROM curves were absent of EMG activity from 20 to 50° in all but one subject. Removing the subject with hamstring EMG activity at 50° did not affect the results (PT SLR ROM vs increase in torque from 20 to 50° r = −0.79, P = 0.0001; PT SLR ROM vs energy absorbed from 20 to 50° r = −0.71, P = 0.003; PT SLR ROM vs increase in torque from 20 to 50° + total energy absorbed r = 0.87, P = 0.0001).


The passive mechanical responses to stretch have been studied extensively in both animals and humans in an attempt to explain musculoskeletal flexibility (7,8,11,12,20,21,22,27). However, none of these studies has attempted to determine the relationship between the mechanical measures and clinical tests of flexibility. The SLR test is an accepted clinical test of hamstring flexibility (7,8,19). The results of this study suggest that the SLR test is primarily a measurement of the passive mechanical forces resisting hip flexion motion. Sixty-six percent of the interindividual variation in PT SLR ROM could be explained by the increase in passive torque from 20 to 50°. The greater the increase in passive torque, the lower the maximum ROM. A further 13% of the variation in PT SLR ROM could be explained by factoring in the total energy absorbed over the total ROM during the SLR stretch. The greater the energy absorbed over the total stretch, the higher the maximum ROM. Torque/ROM curves for two subjects with a large difference in maximum ROM are shown in Fig. 3.

Figure 3
Figure 3:
Torque/ROM curves for a "tight" subject (dotted curve) and a "loose" subject (solid curve). Both subjects were male. Age: "tight" = 25 yr; "loose" = 31 yr. Height: tight = 178 cm; loose = 180 cm. Body mass: tight = 79.5 kg; loose = 77.2 kg. Increase in torque from 20 to 50°: tight = 19.6 N·m; loose = 9.2 N·m. Energy absorbed from 20 to 50°: tight = 7.0 N·m·rad−1; loose = 4.8 N·m·rad−1. Total energy absorbed: tight = 22.7 N·m·rad−1; loose = 50.5 N·m·rad−1.

The lack of a clinically significant stretch-induced EMG response during the instrumented SLR stretch is in agreement with Halbertsma et al. (1994) who used a similar measurement technique (12). It is possible that the greater ROM measured during the clinical SLR test may have elicited a greater stretch-induced EMG response. However, it is more likely that the greater ROM reflected the therapist's limited ability to manually stabilize against pelvic rotation and knee extension and during the measurement. During the instrumented stretch, maximum ROM was determined as the point at which subjects reported discomfort. These findings suggest that perception of discomfort during slow passive stretching precedes any reflex contractile response detectable with surface EMG. It is likely that if the stretch had been continued beyond the point of discomfort, a greater contractile response might have been elicited. However, Davidoff (1992) has suggested that reflex activity only contributes to resistance during rapid stretch or in individuals with certain neurological disorders (4).

Values for passive mechanical restraint to hamstring stretches have been reported previously (7,8,12,21). Halbertsma et al. (1994), using a similar technique to the present study, demonstrated values for passive angular stiffness of approximately 69 N·m·rad−1 from 20 to 40° (12) compared with 5.2 N·m·rad−1 from 20 to 50° in the present study. However, Halbertsma et al. (1994) only included male subjects with clinically "short" hamstrings. Using a different type of hamstring stretch, Gajdosik et al. (1990) reported values for angular stiffness of 40.7 N·m·rad−1 for men and 25.8 N·m·rad−1 for women during submaximal stretch (7). These values compare favorably with the values presented here.

The energy loss evident during unloading of the stretch (57 ± 3%) is similar to values reported for the plantar flexors (51%) (14) but higher than values reported in an animal model (39%) (1). The slightly slower rate of stretch for lowering compared with raising the limb (1.7 ± 1°·s−1 vs 3.1 ± 0.2°·s−1) may have affected the measurement of energy loss.

The mechanical properties of relaxed skeletal muscle have been referred to as the parallel elastic component (PEC), whereas the series elastic component (SEC) refers to the mechanical properties of active skeletal muscle (17,18). Wilson et al. (1991) demonstrated that 30% of the variation in glenohumeral ROM could be explained by stiffness of the SEC (31). In the present study, 79% of the variation in hip flexion ROM could be explained by measurements of the PEC. The capacity of the SEC to contribute to external work performed during human movement is well recognized (2,28). However, the role of the PEC in human movement is less clear. Recent clinical studies have demonstrated a significant relationship between passive flexibility measurements and running economy, with less flexible individuals being more economical (3,9). These effects were attributed to increased storage and return of elastic energy in the less flexible muscles. In the present study, greater energy was absorbed, for a given submaximal ROM during passive stretch, in subjects with a lower maximum ROM. It is possible that the mechanical contribution to external work during human movement reflects the combined effects of the SEC and PEC.


Joint torque measured during passive SLR stretch seemed to be a function of the passive mechanical restraints to joint motion with minimal evidence of a stretch-induced contractile response. Seventy-nine percent of the variability in maximum SLR ROM, a clinical test of hamstring flexibility, could be explained by the passive mechanical restraints to motion. These data lend support to the concept that musculoskeletal flexibility can be explained in mechanical terms rather than by neural theories.


1. Best, T. M., J. McElhaney, W. E. Garrett, and B. S. Myers. Characterization of the passive responses of live skeletal muscle using the quasi-linear theory of viscoelasticity. J. Biomech. 27:413-419, 1994.
2. Cavagna, G. A. Elastic bounce of the body. J. Appl. Physiol. 29:279-282, 1970.
3. Craib, M. W., V. A. Mitchell, K. B. Fields, T. R. Cooper, R. Hopewell, and D. W. Morgan. The association between flexibility and running economy in sub-elite male distance runners. Med. Sci. Sports Exerc. 28:737-743, 1996.
4. Davidoff, D. A. Skeletal muscle tone and the misunderstood stretch reflex. Neurology 42:951-963, 1992.
5. Etnyre, B. R., and E. J. Lee. Comments on proprioceptive neuromuscular facilitation stretching techniques. Res. Q. Exerc. Sci. 58:184-188, 1987.
6. Etnyre, B. R., and L. D. Abraham. Antagonist muscle activity during stretching: a paradox re-assessed. Med. Sci. Sports Exerc. 20:285-289, 1988.
7. Gajdosik, R. L., C. A. Giuliani, and R. W. Bohannon. Passive compliance and length of the hamstring muscles of healthy men and women. Clin. Biomech. 5:23-29, 1990.
8. Gajdosik, R. L. Passive compliance and length of clinically short hamstring muscles of healthy men. Clin. Biomech. 6:239-244, 1991.
9. Gleim, G. W., N. S. Stachenfeld, and J. A. Nicholas. The influence of flexibility on the economy of walking and jogging. J. Orthop. Res. 8:814-823, 1990.
10. Guissard, N., J. Duchateau, and K. Hainaut. Muscle stretching and motoneuron excitability. Eur. J. Appl. Physiol. 58:47-52, 1988.
11. Halbertsma, J. P. K., A. I. van Bolhuis, and L. N. H. Goeken. Sport stretching: effect on passive muscle stiffness of short hamstrings. Arch. Physiol. Med. Rehabil. 77:688-692, 1996.
12. Halbertsma, J. P. K., and L. N. H. Goeken. Stretching exercises: effect of passive extensibility and stiffness in short hamstrings of healthy subjects. Arch. Physiol. Med. Rehabil. 75:976-981, 1994.
13. Hardy, L. Improving active range of hip flexion. Res. Q. Exerc. Sci. 56:111-114, 1985.
14. Hufschmidt A., and K. H. Mauritz. Chronic transformation of muscle in spasticity: a peripheral contribution to increased tone. J. Neurol. Neurosurg. Psychiatry 48:676-685, 1985.
15. Hutton, R. S., and S. W. Atwater. Acute and chronic adaptations of muscle proprioceptors in response to increased use. Sports Med. 14:406-421, 1992.
16. Hutton, R. S. Neuromuscular basis of stretching exercises. In: Strength and Power in Sport, P.V. Komi (Ed.). London: Blackwell Scientific, 1992, pp. 29-38.
17. Jewell, B. R., and D. R. Wilkie. An analysis of the mechanical components in frog's striated muscle. J. Physiol. (Lond.) 143:515-540, 1958.
18. Latash, M. L., and V. M. Zatsiorsky. Joint stiffness: myth or reality? Hum. Movement Sci. 12:653-692, 1993.
19. Liemohn, W. Flexibility/range of motion. In: ACSM Resource Manual for Guidelines for Exercise Testing and Prescription, 2nd Ed. Philadelphia: Lea and Febiger, 1993, pp 327-336.
20. Magnusson S. P., E. B. Simonsen, P. Aagaard, G. W. Gleim, M. P. McHugh, and M. Kjaer. Viscoelastic response to repeated static stretching in the human hamstring muscle. Scand. J. Med. Sci. Sports 5:342-347, 1995.
21. Magnusson, S. P., E. B. Simonsen, P. Aagaard, H. Strensen, and M. Kjaer. A mechanism for altered flexibility in human skeletal muscle. J. Physiol. (Lond.) 497:291-298, 1996.
22. McHugh, M. P., S. P. Magnusson, G. W. Gleim, and J. A. Nicholas. Viscoelastic stress relaxation in human skeletal muscle. Med. Sci. Sports Exerc. 24:1375-1382, 1992.
23. Moore, M. A., and R. S. Hutton. Electromyographic investigation of muscle stretching techniques. Med. Sci. Sports Exerc. 12:322-329, 1980.
24. Osternig, L. R., R. Robertson, R. Troxel, and P. Hansen. Muscle activation during proprioceptive neuromuscular facilitation (PNF) stretching techniques. Am. J. Physiol. Med. 66:298-307, 1987.
25. Osternig, L. R., R. N. Robertson, R. K. Troxel, and P. Hansen. Differential responses to proprioceptive neuromuscular facilitation (PNF) stretch techniques. Med. Sci. Sports Exerc. 22:106-111, 1990.
26. Tanigawa, M. C. Comparison of the hold-relax procedure and passive mobilization on increasing muscle length. Physiol. Ther. 52:725-735, 1972.
27. Taylor, D. C., J. D. Dalton, A. V. Seaber, and W. E. Garrett. Viscoelastic properties of muscle-tendon units: the biomechanical effects of stretching. Am. J. Sports Med. 18:300-309, 1990.
28. Thys, H., G. A. Cavagna, and R. Margaria. The role played by elasticity in an exercise involving movements of small amplitude. Pflugers Arch. 354:281-286, 1975.
29. Watts, R. L., A. W. Wiegner, and R. R. Young. Elastic properties of muscles measured at the elbow in man: II. patients with Parkinsonian rigidity. J. Neurol. Neurosurg. Psychiatry 49:1177-1181, 1986.
30. Wiegner, A. W., and R. L. Watts. Elastic properties of muscles measured at the elbow in man: I. normal controls. J. Neurol. Neurosurg. Psychiatry 49:1171-1176, 1986.
31. Wilson, G. J., G. A. Wood, and B. C. Elliott. The relationship between the stiffness of the musculature and static flexibility: an alternative explanation for the occurrence of muscle injury. Int. J. Sports Med. 12:403-407, 1991.


© Williams & Wilkins 1998. All Rights Reserved.