Localized muscle fatigue is often defined as an acute impairment of performance that includes both an increase in the perceived effort necessary to exert a desired force and an eventual inability to produce this force (4,11). While most research on muscle fatigue has focused on changes in contractile properties of muscles, less attention has been devoted to the effects on proprioception or integrative motor coordination (for review see (13)). This is somewhat surprising since fatigue is commonly associated with impaired motor performance (4,11). In this context the effects of muscle fatigue on proprioception are potentially interesting since proprioceptive afferent feedback is a crucial factor for optimal motor control (36,37). The current consensus is that the sense of proprioception originates in the simultaneous activity of a range of different types of receptors located in muscles, joints, and skin (for review see (15)). Some of these receptors have been shown in animal studies to be affected by muscle fatigue (17,20,26,31,32) and/or by increased intramuscular concentrations of substances released during muscle contractions (8-10,21,33). Although it is reasonable to assume that these receptors are affected in a similar way in humans, comparably little is known about the fatigue effects on human proprioception.
Since the definition of proprioception in the literature is equivocal, in this paper it is defined as the movement and position sense and will be considered to be synonymous with the ability to detect, without visual input, the spatial position and/or movement of limbs in relation to the rest of the body. A few studies, using various paradigms investigating fatigue effects on proprioception, have yielded somewhat conflicting results (28,38,39,41). In the only study done on the shoulder, Voight et al. (41) reported that the "position sense" is diminished in the presence of shoulder muscle fatigue. However, that study (41) and others (30) have not investigated the influences of fatigue on the "movement sense," which is recognized to be a separate submodality of proprioception.
The shoulder joint is very complex and includes several bone and joint structures. The functional stability is maintained through the collaborative effect of ligaments and the rotator cuff muscular complex, as well as other muscles. This makes shoulder movement schemes very complex and the contribution to the proprioceptive inflow from different receptors in the different joint structures hard to establish. Therefore, in our study we have chosen not to focus on any particular muscle or joint structure, but to study the ability to differentiate between movement velocities for passive movements imposed over the shoulder joint. Most studies on proprioception carried out on the shoulder have focused on the ability to match or reproduce joint positions or on movement detection thresholds (e.g. 3,5,25,41). The present study used a novel approach focusing exclusively on the movement sense.
The aim of the present study was to study effects of muscle fatigue on the acuity of the movement sense. To this end we measured the ability to discriminate between different velocities of imposed rotations in the shoulder joint before and after induction of localized muscle fatigue.
Subjects. This study was conducted on 14 subjects (8 male and 6 female) with a mean age 23 ± 2 yr. Eleven subjects had regular physical training three or more times per week while two trained less than three times a week and one had no regular training. The average height was 184.4 ± 7.4 cm for men and 172.8 ± 6.8 for women. The average weight was 82.9 ± 11.1 kg and 62.1 ± 7.2 kg for the men and women, respectively. All subjects were right handed and without a history of musculoskeletal problems in the upper body.
The study was approved by the ethical committee at the Faculty of Medicine at the University of Umeå and the subjects provided written informed consent.
Apparatus for measuring movement sense. The apparatus for measurement of movement sense is illustrated in Figure 1A. It consisted of a steady chair and a rig shaped to fit the subject's right forearm. The rig was controlled for rotations in the horizontal plane by a PC-controlled servomotor. In addition, the angular velocity and range of rotation of the rig, the entire test sequence for an experiment (see below) could also be preprogrammed. The rig and the chair were adjustable so that the arm of each subject was always placed with 65° flexion, 30° abduction, and a slight inward rotation in the shoulder joint. A 30° flexion in the elbow joint and the lower arm at 80° of pronation (denoted starting position below) was also standardized for each subject. This configuration was chosen because it is comfortable and can be well tolerated for long periods even by patients with pain in the shoulder-neck area. The adjustment was also made so that the rotation axis of the rig was congruent with the rotation center of the glenohumeral joint. Therefore, rotations by the motor driven rig could impose passive rotations in the glenohumeral joint in the transversal plane of the subject.
The angle of the rig was recorded using an electromagnetic tracking system which had an angular resolution of 0.01° (FASTRAK System, Polhemus Inc., Colchester, VT). The FASTRAK electromagnetic tracker system uses a magnetic receiver relative to a fixed magnetic transmitter. The transmitter generates a weak magnetic field, the receiver registers the induced changes in amplitude and phase displacements in the magnetic field, and the main unit calculates six orientation values (x-, y-, z-position and azimuth, elevation, and roll angle) for the receiver. The system's electromagnetic transmitter was placed just below the rig and the receiver was attached beneath the rig (Figs. 1A and B). The angle of the rig was sampled at 120·s−1 and stored on computer file. The angular velocity of the imposed rotations was calculated off-line.
Experimental procedure. Subjects were seated in the apparatus while wearing blindfolds and headphones emitting white noise to minimize visual or auditory cues, e.g., noise from the motor. The range of the imposed rotations is illustrated in Figure 1B. The starting position for the movements was always 85° (relative to the frontal plane) and the end position varied from 40 to 20°.
The subjects were instructed to produce a light torque in the opposite direction of the movement a few seconds before and throughout the rotations. During an initial training period subjects were provided feedback on the torque to keep it at 5 Nm. After each movement subjects actively returned the rig to the starting position.
Each subject was tested in two experiments denoted "Light exercise" and "Hard exercise," differing only with respect to the conditioning paradigm used (see below). The order of the two experiments was randomized for each subject and separated by at least 1 day.
In each experiment the subjects were presented with 60 trials. Each trial consisted of a reference velocity, i.e., an imposed rotation of the glenohumeral joint, at 50°·s−1 followed by a test velocity that was within the 40-60°·s−1 range. The test velocities were never equal to the reference velocity but otherwise were normally distributed around the reference velocity. With the use of the pseudorandomizing ABBA method (35), it was ensured that the faster and slower test velocities were equally distributed over the experiment; in all other aspects the order of the 60 test velocities were totally randomized. The time and range of the velocity presentations (i.e., the movements) were also randomized to minimize any clues from either of these entities. This is illustrated by the shaded range between 40° and 20° at the end of the movement in Figure 1B. To minimize any influence from the experimentalist, all instructions given during the experiments were prerecorded on computer files and admitted via subject headphones.
The tests were done according to the forced choice paradigm, i.e., for each trial the subjects had only two response alternatives. They were instructed to report if they perceived the test velocity as either faster or slower than the reference velocity. At the beginning of the two experiments the subjects were instructed to trust their intuition if they felt insecure and give an answer accordingly. If the subjects did not answer within 3 s the trial was discarded.
At the beginning of the experiments the subjects were given a brief training period to acquaint them with the experimental procedure. The experiments were divided into four test periods of 15 trials each. Before and in between these periods the subjects were conditioned with a preset exercise task.
Conditioning exercise task. The subjects were seated and strapped in an isokinetic dynamometer, i.e. a Cybex machine (Cybex II, Lumex Inc., Long Island, NY). By connecting the Cybex machine to an oscilloscope we were able to continuously monitor the forces produced during the exercise. The exercise task consisted of repetitive isokinetic horizontal extension-flexion rotations of the right shoulder at 30°·s−1 over the same range of motion as in the tests (see Fig. 1). The machine was adjusted so that the angles in the elbow and shoulder joints were similar to the test situation.
Before the first test period the subjects performed a short warm-up exercise. Thereafter the subjects were instructed to do three consecutive pairs of horizontal extensions and flexions at maximal voluntary contraction (MVC). The peak torques for both directions were registered. After 1 min of rest the subjects performed three 2-min exercise sessions with 1-min pauses in between. The exercise in these sessions consisted of repetitive horizontal extensions and flexions. In the Hard exercise experiment the subjects worked at their maximal capacity throughout the sessions. If the subjects produced more than 30% of the initial MVC at the end of the last session, it was continued until the torque fell below 30% of MVC. In between the other test periods the subjects performed a exercise task according to the rules of the last exercise session. In the Light exercise experiments the subjects performed the same exercise task except that they were never allowed to exceed 10% of initial MVC. Thus, the experiments were identical apart from the workload.
Analysis. To statistically investigate the influence of any of the factors, type of conditioning (i.e., Hard exercise or Light exercise), negative or positive velocity difference, and subject gender on the probability for a correct answer, a model based on a logistic regression analysis was made from the pooled responses from all subjects to the 60 trials in each experiment. The model was built on the assumption that for both conditions the probability of a correct answer should be 0.5 for the close to zero velocity difference and close to 1.0 for the large differences in velocity.
The equation of the model is: (Equation 1)
In Equation 1, H equals the absolute value of the velocity difference. The absolute value was used since the probability of a correct answer would increase regardless of increasing negative or positive velocity difference. A represents Hard exercise or Light exercise condition, S indicates negative or positive velocity difference, and K the subject gender. The function for estimation of the probability of a correct answer (G = 1) in the logistic regression is given in Equation 2.
The statistical tests were based on the ratio between the estimate of a parameter in the model and its estimated SE. Significance was assessed by comparing the observed ratio with the standard normal distribution. The analysis was based on 1675 observations for the 14 subjects.
Figure 2 illustrates the result for one male subject from the Light exercise (Fig. 2A) and the Hard exercise (Fig. 2B) experiments. In the diagrams a vertical filled bar represents a correct answer on a trial. The 60 different trials are equidistantly distributed along the x-axis and ordered from negative to positive test-reference velocity. It can be determined from the figure that during the Light exercise experiment, the subject gave quite accurate answers for the greater absolute velocity differences and that the number of wrong judgments increased as the test-reference velocity difference approached zero. The accuracy for the greater velocity differences was not as obvious in the Hard exercise experiment (Fig. 2B); especially for the positive velocity differences the subject appeared less accurate than in Light exercise. For this subject there was a significantly higher probability of a correct answer during Light exercise compared with Hard exercise (P < 0.01).
Figure 3 shows the results from the logistic regression analysis of the pooled data from the Light exercise (thin lines) and the Hard exercise (bold lines) experiments from all male (dashed lines) and female (filled lines) subjects. The lines represent the estimated probability of a correct answer as a function of the test velocity-reference velocity differences. There is a consistently lower probability of a correct answer during the fatiguing Hard exercise experiments compared with the nonfatiguing Light exercise experiments, and this difference was statistically significant (P < 0.001). There was also a significant difference between men and women (P < 0.001).
Finally, when the test velocity was slower than the reference velocity there was a higher probability of a correct answer than when the test velocity was faster than the reference velocity (P < 0.001).
This study investigated the effects of shoulder muscle fatigue on the acuity of the movement sense in the same shoulder, and the results were threefold: 1) There was a clear reduction in the acuity of the movement sense during localized muscle fatigue; 2) this reduction was greater when the test velocity was faster (as opposed to slower) than the reference velocity, and 3) the women had a lower acuity in the movement sense than men. These findings were presented in part in a previous abstract (18).
For movement sense and position sense, acuity of the shoulder shows similar effects of localized muscle fatigue. Thus, Voight et al. (41) reported a decline in the ability to actively or passively reproduce positions. Similarly, Sharpe and Miles (38) found a decline in the ability to reproduce elbow positions after localized muscle fatigue. Since they observed the same effect with the nonfatigued contralateral elbow used as control, they suggested that the decline may have resulted from central instead of localized fatigue. However, it has been shown in several animal studies that increased intramuscular concentrations of contraction metabolites and inflammatory substances liberated during muscle contractions in both contralateral and heteronymous muscles strongly influence the γ-muscle-spindle system via chemosensitive muscle afferents (8-10,19,21,27,33). Note that the muscle spindle system is known to be involved in proprioception. In humans vibration of the biceps brachii or triceps brachii has been shown to alter the position sense in the contralateral arm (24). Therefore, it is difficult to draw any conclusion on central or peripheral effects from the study by Sharpe and Miles (38). In another study (39) it was concluded that localized muscle fatigue caused a decrease in the ability to reproduce knee joint angles.
An important issue is whether the effects on movement sense observed in the present study can be attributed to central fatigue or to muscle fatigue. Obviously, central fatigue (e.g., overuse of central pathways or networks) may have accompanied peripherally elicited effects, but there is a chain of evidence indicating that alterations in the proprioceptive inflow from peripheral muscle receptors have contributed considerably to the observed effects. First, the conditioning isokinetic exercise undoubtedly produced a muscle fatigue in both antagonist and agonist muscles, and, as briefly described in the introductory section, several types of muscle receptors are affected by localized muscle fatigue (17,20,26,32). Second, the localized muscle fatigue probably increased the intramuscular concentrations of several contraction metabolites and inflammatory substances liberated during muscle contractions (e.g., lactic acid, KCl, bradykinin, arachidonic acid, serotonin). Increased concentrations of those substances are known to alter the muscle spindle output via reflex effects on the fusimotor system from chemosensitive muscle afferents (8-10,21,23,33). Since the muscle spindles are strongly affected by the aforementioned substances and by localized muscle fatigue (26,27), the proprioceptive inflow from spindle afferents during the tests is likely to have been affected by the conditioning procedure. Third, it is known from earlier studies that a small voluntary contraction resisting an imposed movement increases the afferent output from the muscles, mainly from the muscle spindle afferents through an increased fusimotor drive (2,12,16,29). Since our subjects were instructed to produce a small torque in opposition to the imposed movements, this is likely to have further increased the contribution of the muscle spindles to the proprioceptive inflow. Fourth, the forced choice experimental design has been shown to be relatively insensitive to cortical and/or conscious influences (6,34,40). Therefore, it could be argued that effects on the ability to discriminate correctly between movement velocities from central fatigue on these levels was minimized by using this paradigm. Even though no definite conclusions can be made concerning the issue of whether the observed reduction in the acuity of the movement sense was caused by mainly central or peripheral effects, the above line of evidence indicates that the afferent output from the muscle receptors and especially from the muscle spindle afferents have influenced the results.
Obviously, disturbances in the proprioceptive input from muscle receptors, and especially from the muscle spindle system, would imply effects on motor control (36,37) and dexterity. This may have several implications, among them a decrease in the performance and an increased risk for injuries during sports activities or during hard exercise. Furthermore, in the present study the movement sense acuity was lower for the women. These observations have to be interpreted with caution, but if generalized they might imply that women are more prone to sport injuries than men. Interestingly, some studies lend some support to such speculations, as they suggest that women are more susceptible to sport injuries than men (1,7,14,22).
1. Arendt, E. and R. Dick. Knee injury patterns among men and women in collegiate basketball and soccer: NCAA data and review of literature. Am. J. Sports Med.
2. Barrack, R. L., H. B. Skinner, M. E. Brunet, and S. D. Cook. Joint laxity and proprioception
in the knee. Physician Sportsmed.
3. Blasier, R. B., J. E. Carpenter, and L. J. Huston,. Shoulder proprioception
: effect of joint laxity, joint position, and direction of motion. Orthop. Rev.
4. Chaffin, D. B. Localized muscle fatigue: definition and measurement. J. Occup. Med.
5. Cohen, L. A. Analysis of position sense in human shoulder. J. Neurophysiol.
6. Cowey, A. and P. Stoerig. Blindsight in monkeys. Nature
7. De Loes, M. Epidemiology of sports injuries in the Swiss organization "Youth and Sports" 1987-1989: injuries, exposure and risks of main diagnoses. Int. J. Sports Med.
8. Djupsjöbacka, M., H. Johansson, and M. Bergenheim. Influences on the γ-muscle spindle system from muscle afferents stimulated by increased intramuscular concentrations of arachidonic acid. Brain Res.
9. Djupsjöbacka, M., H. Johansson, M. Bergenheim, and P. Sjölander. Influences on the γ-muscle-spindle system from contralateral muscle afferents stimulated by KCl and lactic acid. Neurosci. Res.
10. Djupsjöbacka, M., H. Johansson, M. Bergenheim, and B. I. Wenngren. Influences on the γ-muscle spindle system from muscle afferents stimulated by increased intramuscular concentrations of bradykinin and 5-HT. Neurosci. Res.
11. Enoka, R. M. and D. G. Stuart. Neurobiology of muscle fatigue. J. Appl. Physiol.
12. Ferrell, W. R. and A. Smith. The effect of loading on position sense at the proximal interphalangeal joint of the human index finger. J. Physiol.
13. Gandevia, S. C., R. M. Enoka, A. J. McComas, D. G. Stuart, and C. K. Thomas. (Eds.). Neurobiology of muscle fatigue, advances and issues. In: Fatigue, Neural and Muscular Mechanism,
New York: Plenum Press, 1995, pp. 515-524.
14. Gray, J., J. E. Taunton, D. C. McKenzie, D. B. Clement, J. P. McConkey, and R. G. Davidson. A survey of injuries to the anterior cruciate ligament of the knee in female basketball players. Int. J. Sports Med.
15. Grigg, P. Peripheral neural mechanisms in proprioception
. J. Sport Rehabil.
16. Grigg, P., G. A. Finerman, and L. H. Riley. Joint-position sense after total hip replacement. J. Bone Joint Surg.
55-A: 1016-1025, 1973.
17. Hayward, L., U. Wesselmann, and W. Z. Rymer. Effects of muscle fatigue on mechanically sensitive afferents of slow conduction velocity in the cat triceps surae. J. Neurophysiol.
18. Hellström, F., J. Lönn, J. Pedersen, M. Djupsjöbacka, and H. Johansson. Localized muscle fatigue reduces movement sense acuity in females to a greater extent than males (Abstract). Med. Sci. Sports Exerc.
19. Hong, S. K., K.-D. Kniffki, and R. F. Schmidt. Effects of intraarterial injection of algesic agents into skeletal muscle on the discharges of homonymous extensor gamma motoneurones. Pflügers Arch.
20. Hutton, R. S. and D. L. Nelson. Stretch sensitivity of Golgi tendon organs in fatigued gastrocnemius muscle. Med. Sci. Sports Exerc.
21. Johansson, H., M. Djupsjöbacka, and P. Sjölander. Influences on the γ-muscle spindle system from muscle afferents stimulated by KCl and lactic acid. Neurosci. Res.
22. Jones, B. H., M. W. Bovee, J. M. D. Harris, and D. N. Cowan. Intrinsic risk factors for exercise-related injuries among male and female army trainees. Am. J. Sports Med.
23. Jovanovic, K., R. Anastasijevic, and J. Vuco. Reflex effects on fusimotor neurones on chemically induced discharges in small diameter muscle afferents in decerebrate cats. Brain Res.
24. Lackner, J. R. Some influences of tonic vibration reflexes on the position sense of the contralateral limb. Exp. Neurol.
25. Lephart, S. M., J. J. P. Warner, P. A. Borsa, and F. H. Fu. Proprioception
of the shoulder joint in healthy, unstable, and surgically repaired shoulders. J. Shoulder Elbow Surg. Bd. Trust.
26. Ljubisavljevic, M. and R. Anastasijevic. Fusimotor-induced changes in muscle spindle outflow and responsiveness in muscle fatigue in decerebrate cats. Neuroscience
27. Ljubisavljevic, M., K. Jovanovic, and R. Anastasijevic. Changes in discharge rate of fusimotor neurones provoked by fatiguing contractions of cat triceps surae muscles. J. Physiol. (Lond.)
28. Marks, R. and H. A. Quinney. Effect of fatiguing maximal isokinetic quadriceps contractions on ability to estimate knee-position. Percept. Mot. Skills
29. Marks, R., H. A. Quinney, and J. Wessel. Proprioceptive sensibility in women with normal and osteoarthritic knee joints. Clin. Rheumatol.
30. McCloskey, D. I. Differences between the senses of movement and position shown by the effects of loading and vibration of muscles in man. Brain Res.
31. Nelson, D. L. and R. S. Hutton. Dynamic and static stretch responses in muscle spindle receptors in fatigued muscle. Med. Sci. Sports Exerc.
32. Pedersen, J., M. Ljubisavljevic, M. Bergeneheim, and H. Johansson. Alterations in information transmission in ensembles of primary muscle spindle afferents after fatigue in heteronymous muscle. Neuroscience
33. Pedersen, J., P. Sjölander, B.-I. Wenngren, and H. Johansson. Increased intramuscular concentration of bradykinin increases the static fusimotor drive to muscle spindles in neck muscles of the cat. Pain
34. Perenin, M. T., J. Ruel, and H. Hecaen. Residual visual capacities in a case of cortical blindness. Cortex
35. Riggs, L. and J. Kling. Introduction. In: Woodworth & Schlosberg's Experimental Psychology.
London: Methuen & Co., 1971, p. 5.
36. Sainburg, R. L., H. Poizner, and C. Ghez. Loss of proprioception
produces deficits in interjoint coordination. J. Neurophysiol.
37. Sanes, J. N., K.-H. Mauritz, M. C. Dalakas, and E. V. Evarts. Motor control in humans with large-fiber sensory neuropathy. Hum. Neurobiol.
38. Sharpe, M. H. and T. S. Miles. Position sense at the elbow after fatiguing contractions. Exp. Brain Res.
39. Skinner, H. B., M. P. Wyatt, J. A. Hodgdon, D. W. Conard, and R. L. Barrack. Effect of fatigue on joint position sense of the knee. J. Orthop. Res.
40. Weiskrantz, L. Outlooks for blindsight: explicit methodologies for implicit processes. Proc. R. Soc. Lond. B Biol. Sci.
239:247-278, 1990, 1989.
41. Voight, M. L., J. A. Hardin, T. A. Blackburn, S. Tippett, and G. C. Canner. The effects of muscle fatigue on and the relationship of arm dominance to shoulder proprioception
. J. Orthop. Sports Phys. Ther.