Abstract
Proprioceptive mechanisms appear to play a role in stabilizing the joints and may serve as a means for interplay between static stabilizers and dynamic muscular restraints. The purpose of our study was to investigate whether or not gymnastic training has any effect on the balance and on proprioception in an ankle, as seen in gymnasts and in nongymnasts. We evaluated the proprioceptive ability of the ankle using four different tests (a one-leg-standing test, a single-limb-hopping test, an active angle-reproduction test, and a passive angle-reproduction test). Proprioception of the ankle was measured in 40 subjects who were assigned to two experimental groups. Group 1 (n: 20) were healthy control subjects, and group 2 (n: 20) were teenaged female gymnasts.
The sense of position of a joint was actively measured using a Cybex NORMTM isokinetic dynamometer and measured passively with a proprioception-testing device. A Mann-Whitney U test was used to compare mean values of the gymnasts to the controls. Results revealed statistically significant differences (p<0.05) between two groups. We found no statistically significant differences between the dominant and nondominant ankle in volunteers or in gymnasts, in all tests. The results of this study suggest that gymnastic training has a positive influence on sense of position of the ankle joint and on balance, in addition to increasing muscle tone.
INTRODUCTION
Over the last 10 years, there has been greater emphasis on research in proprioception, focusing on various anatomical areas, but most commonly on the knee and the ankle. Proprioception is the cumulative neural input to the central neural system from mechanoreceptors located in the joint-capsule, ligaments, muscles, tendons, and skin. 20 Articular mechanoreceptors in the ankle, 22 knee, 15,16,24 and shoulder 21,27 have been identified morphohistologically in both animal 9,17 and human models, suggesting an anatomical basis for an active proprioceptive mechanism in all joints. The proprioceptive mechanism is essential for proper function of the joint in sports, for activities of daily living, and for some occupational tasks.
It is widely known that a proprioceptive deficit may detract from the functional success of healing in ligaments and may predispose the patient to reinjury. Thus, assessment of proprioceptive sensibility is valuable for identification of proprioceptive deficits and for subsequent planning of a rehabilitative program. If we could enhance proprioception of a joint, we might be able to restore the normal protective mechanism within the injured or reconstructed joint. Rehabilitation must therefore focus on restoring the proprioceptive mechanism by enhancing cognitive appreciation of the respective joint relative to position and movement, and by providing muscular stabilization of the joint in the absence of structural restraints. 20
Although their results are inconsistent, a few available studies have shown that gymnastic training influences proprioception of the knee 2,19 and ankle. 4,12,23 In addition, we have observed that gymnasts have better balance than nonathletes. This is the first study to evaluate the ability to sense the position of the ankle-joint and balance in gymnasts. No studies, however, have reported the effect of gymnastic training on balance and on proprioception of the ankle of gymnasts. Gymnasts were chosen as the group to be studied because they combine muscular development and flexibility with a constant awareness of position and motion of the joint. 19
The purpose of our study was to investigate whether or not gymnastic training has any effect on the balance and on proprioception of the ankle of gymnasts. We wanted to know what effect, if any, extensive athletic training has on the active and passive sense of position of the ankle and on balance. This study was designed to provide objective information relative to the status of passive and active sense of the position of the ankle and of balance on the ankle of a gymnast. Our hypotheses were twofold:
Gymnasts, given their better balance, also have greater balancing and proprioceptive ability than healthy nongymnasts;
Limb dominance in both gymnasts and controls does not influence proprioceptive sensibility and balance.
METHODS
Subjects and Criteria for Inclusion
A total of 40 subjects participated in these studies between February 1998, and October 1999 at the Department of Sports Medicine of Gülhane Academy of Military Medicine. We designed the study to compare a group of healthy teenaged gymnasts with a control group of healthy age-matched volunteers. All subjects were volunteers, met the criteria for inclusion, and provided informed consent as approved by the Human Subjects Review Board of our institution. Two experimental groups were formed. The criteria for inclusion for both groups were as follows: age 10 to 17 years, no history of injury in either ankle, knee, or hip, no musculoskeletal injury, no abnormality of the inner ear, no disorder of equilibrium, and no neurological disease. Each prospective subject completed a questionnaire documenting the criteria for inclusion, gave information about general health, and gave demographic data. Prior to participation in the study, the subjects and the parents or guardians of those under 18 years of age were required to give informed consent that met the requirements of a local human subjects institutional review board. Table 1 presents descriptive data on the subjects.
Physical characteristics of subjects (n: 40)
Groups Studies
Group 1 (n=20) included 20 healthy, nongymnastic volunteers. All subjects considered themselves athletically active but did not regularly participate in any sports involving the lower extremity (i.e., football, running). Group 2 (n=20) included 20 healthy teenaged, female gymnasts. They were requested to refrain from unusual activities or vigorous exercise 24 hours before their testing session. These gymnasts were tested at least three years after beginning gymnastic teaching (mean 7.3 years±SD 2.7 years).
Testing Procedures
We evaluated the proprioceptive ability of the ankle using four different tests. We used a one-leg-standing test, a single-limb-hopping test, an active angle-reproduction test, and a passive angle-reproduction test. The one-leg-standing test and the single-limb-hopping course incorporated the principles discussed by Jerosch and colleagues. 14 In an attempt to minimize the effect of fatigue on the testing, the passive sense of position of a joint was performed first, followed by testing of active reposition, the one-leg-standing test, and the single-limb-hopping course. A practice session was immediately followed by the testing session.
Single-Limb-Hopping Course
This test is especially useful in documenting the function of the ankle on an uneven surface. 14 The jumping course consists of eight squares, four of which are flat. One square has a 15° incline. Another square has a 15° decline, and two squares show a 15° lateral inclination (Fig. 1). The volunteers are asked to jump across this course on one leg by touching each area once as fast as possible without leaving the course (Fig 2). The test result is quantified by the seconds used to pass the course. Each failure adds an extra second to the time taken to complete the course.
One-Leg-Standing Test
The one-leg-standing test evaluates the volunteer's ability to keep balance while standing on one leg. 1,6,14,23 The volunteer is asked to stand on one leg for one minute with the eyes open and for another minute with the eyes closed to exclude visual perception. In contrast to the current literature we performed this test not on a hard surface but on a soft surface, to increase the failure rate (Fig. 3). Each surface-contact with the contralateral leg was counted as one failure point.
Single-limb hopping course.
Set-up of the single-limb hopping course.
The Active Angle-Reproduction Test
For the test of active reproduction of an angle, a Cybex Norm dynamometer served as the position-sense testing device. The reliability and validity of this device have been favorably demonstrated in several studies. 11,14,26 Tests that employ active reproduction of passive positioning have been used frequently and are accepted tests of proprioception. 10,11,14 Testing was performed at positions of 30° of plantarflexion (Fig. 4) and 15° of inversion (Fig. 5). The tested foot was placed on the plantar flexion-dorsiflexion footplate of the Cybex, according to the manufacturer's instructions for isolating inversion-eversion and plantar flexion-dorsiflexion, and was secured with Velcro straps. 5 For this study, the dominant ankle served as the testing limb for all tests since proprioception of the lower extremity does not appear to be influenced by limb-dominance. 19,21 Prior to testing, the Cybex dynamometer was calibrated as part of the regular schedule for maintenance of equipment used for this testing device. 5
Subject performing one-leg standing test.
Position of angle-reproduction test at Cybex NORM (30° plantarflexion).
To initiate the test, the foot was placed in the neutral (0°) position. All subjects were blindfolded in an effort to eliminate the contribution of visual cues to repositioning of the joint. To familiarize themselves with the testing device, subjects were instructed to perform three active repetitions of ankle movement ranging from maximal plantarflexion to maximal dorsiflexion. The test began with the tester passively moving the tested limb into the testing position of 30° of plantarflexion and maintaining that position for 10 seconds. After 10 seconds of static positioning, the ankle was moved back passively from the presented angle to the reference angle. The subject was asked to actively reproduce the previously presented test angle of 30° of plantarflexion. Two trials were performed. Following the first test, the same testing protocol was used for the second test of active reproduction of passive positioning at 15° of inversion. Angular displacement was recorded as the error in degrees between the presented angle and the repositioned angle. The mean of the two trials for each tested condition was calculated, to determine an average error in scores.
Set-up of angle-reproduction test at the Cybex NORM (15° inversion).
Evaluation of Sense of the Passive Movement
Data were collected using the instrumentation and procedures developed and described by Lentell. 18 The device is a box with a movable platform that rotates about a single axis. This platform is moved by an electric motor that rotates the foot on an axis at a rate of 0.5°/sec. Movement can be stopped at any time by a hand-held switch. The angular motion achieved by the platform is measured by an attached stationary protractor, with measurements rounded off to the nearest degree by visual sighting of the static placement of the parallel surface of the platform (Fig. 6).
Proprioception testing device.
To further reduce unwanted sensory input, subjects were blindfolded and wore a headset with music playing to eliminate the sight and sound of the apparatus moving the foot. The same testing protocol was used for the passive reproduction of passive positioning at 30° of plantarflexion and 15° of inversion. Twelve of the subjects were selected randomly for a retest in five days. They were retested in the same order of conditions and by the same investigator as during the regular testing sessions. The test-retest measures were used to determine the reliability of the testing protocol. Preliminary values on 12 patients revealed a test-retest reliability of r=0.84 for proprioceptive testing.
Statistical Analysis
A Mann-Whitney U test was used to compare the mean values for gymnastic ankles to means for ankles of control. The level of significance for all statistical analysis was set at a p value of <0.05.
RESULTS
Results revealed statistically significant differences between the trained gymnast group and the untrained control group (Table 2).
Comparison between the trained gymnastic group and the untrained control group ankle joints of all four tests.
Single-Limb-Hopping Course
The control and gymnastic groups performed the course two times with each leg. The mean value of the two scores was taken as the total score. Test scores ranged from 6.5 to 12 seconds for the control group, and from 5 to 7.5 seconds for the gymnastic group. The average score in the control group was 9.35±1.6 seconds, while the average score in the gymnastic group was 6.23±0.8 seconds. We found no statistically significant difference between the dominant and nondominant ankle in the group of volunteers or in gymnasts.
One-Leg-Standing Test
The failure rate (ground contact with the contralateral leg) for the one-leg-standing test ranged between two and 21 failure points for the control group and 0.5 and 9 failure points for the gymnasts. The average failure for controls was 9.75±6.0 and for gymnasts 3.0±2.0. There was no statistically significant difference between the dominant and nondominant ankle in volunteers or in gymnasts.
Passive for Sense of Position of the Joint
A Mann-Whitney U test revealed no significant differences in reproduction of passive positioning between dominant and nondominant ankles in either group for any of the tested conditions. The mean values for reproduction of passive positioning ranged from 0.78° to 1.68° for gymnasts and 3.25° to 4.78° for controls. The mean value of scores in degrees of proprioception of a normal ankle are shown in Table 2.
Active for Sense of Position of the Joint
No significant differences in mean scores were revealed between dominant and nondominant ankles in either group for reproduction of active positioning in any of the tested conditions. Mean values for RAP ranged from 3.45° to 3.75° for controls and 1.23° to 2.70° for gymnasts (Table 2). Comparison of ankle joints of the group of trained gymnasts to ankles of the untrained group of controls showed highly significant differences (p<0.05) in all tested conditions (Table 2). The time used for the single-limb-hopping course was 6.23±0.8 seconds for trained gymnasts and 9.35±1.6 seconds for untrained controls (U=12.5; p=0.0001). Scores of the one-leg-standing test averaged 3.0±2.0 points for trained gymnasts and 9.75±6.0 points for untrained controls (U=46; p=0.0001).
The group of untrained controls demonstrated significantly less accurate values for reproduction of passive positioning than did gymnasts, when tested from a starting position of 0° neutral to plantarflexion of 30° (4.78±4.5 compared to 1.68±0.7; U=124; p=0 .037) and to inversion of 15° (3.25±1.9 compared to 0.78±0.7; U=38.5; p=0.0001). The group of gymnasts had significantly lower values than controls for reproduction of active positioning, when tested from a starting position of 0° neutral to plantarflexion of 30° (3.75±2.8 compared to 2.70±2.6; U=130; p=0.041) and to inversion of 15° (3.45±1.3 compared to 1.23±0.6; U=24; p=0.0001). Gymnasts were 33.4% faster than controls in completing the single-limb-hopping course. With a lower failure rate for the one-leg-standing test, gymnasts were 69.2% more successful than controls. Gymnasts made fewer errors than controls in passively reproducing position of the ankle (76% for inversion and 65% for plantarflexion), and fewer errors in actively reproducing position of the ankle (64.3% for inversion and 28% for plantarflexion).
DISCUSSION
The results of our study indicate that gymnasts consistently had better proprioceptive ability than nongymnasts in all tests (Table 2). According to the findings of this and other studies, 2,12,19 highly trained athletes demonstrate a significantly less accurate reproduction of passive positioning and of active positioning and better balance during the test of one-legged-stance and the single-limb-hopping course. This implies enhanced neurosensory pathways. It is possible that athletes are able to develop enhanced neurosensory pathways as a result of long-term athletic training.
Sense of Joint Position
We found that gymnasts were markedly better able to reproduce passive positioning of the ankle joint and active positioning of angles of the ankle for both motions of inversion and plantar flexion as compared to the ability of controls.
Our study revealed that the sense of passive position was significantly better than the sense of active positioning sense in both groups. We found that the maximum inversion and plantarflexion position had the lowest mean score when tested passively and the highest mean score when tested actively. These are consistent with the findings of Bernier and Perrin. 4 Our results are not consistent with the findings of Glencross and Thornton, 8 who found greater error in reproduction of angles for joint position with the largest angles of movement. As their tested angle approached the limit of range of motion, the error in reproduction became greater.
When tested at a slow angular velocity (0.5 to 2° per second), the threshold for detection of passive motion as well as for the reproduction of passive positioning may selectively stimulate Ruffini- or Golgi-type mechanoreceptors. Because the test is performed passively, it may maximally stimulate joint receptors, thereby relying on the cortical pathway in the neuromuscular control system. Stimulation of both joint and muscle receptors is done by the reproduction of active positioning, which provides a more functional assessment of the afferent pathways. Muscle receptors may be more involved in the perception of movement of a joint and may be less valuable in fine judgments of position of a joint than are receptors in the joint. This model may be used to understand better why the total error for active judgments of joint position was significantly greater than the total error for passive judgments.
One-Leg-Standing Test and Single-Limb-Hopping Course
When compared to the control group, trained gymnasts were markedly increased in their ability to balance.
Proprioceptive deficits have been shown to exist following ankle injuries. 1,5,10,14,18,23 Jerosch et al. 14 examined the proprioceptive capabilities of the ankle in stable and unstable joints. Angle reproduction, hopping on one leg, and standing on a soft surface were performed. They found that all the tests showed significant differences between injured and noninjured ankle joints. Gleitz et al. 7 documented one-leg-standing ability in injured and noninjured ankles. They too found a proprioceptive deficit on this test.
Several studies have demonstrated that proprioceptive training involving the ankle disk decreases injury-induced changes in postural sway. 4,12,23 Gleitz et al. 7 and Freeman et al. 6 found an increase in proprioceptive function after neurophysiological training. Gymnasts use both static and dynamic balancing abilities in their daily routines. Therefore, our results on the static one-legged test of balance may not be reflective of dynamic abilities to balance in activities performed routinely by gymnasts. Winstein has determined that tasks testing static balance are not representative of tasks that test dynamic balance. 28 Assessments of abilities to balance, therefore, should be performed in the context in which the subject will be using those abilities. Winstein found that abilities to balance are task-specific; thus, female gymnasts should be tested using activities on the balance beam or other activities specific to their training.
There are some possible explanations for the superior sense of position of a joint and superior balance found in gymnasts. Gymnasts are able to develop enhanced neuromuscular control (NMC) and functional stability as a result of long-term athletic training. Neuromuscular control is influenced by proprioceptive, kinesthetic, visual, and vestibular information as well as cortical and spinal motor commands. 25 Four elements crucial for re-establishing neuromuscular control and functional stability are: joint proprioception and kinesthesia, dynamic stability, preparatory and reactive muscle characteristics, and conscious and unconscious functional motor patterns. 25
Athletes who inherently possess enhanced joint proprioception may excel at sports requiring high levels of neuromuscular control. To improve NMC and functional stability, athletes may use techniques for exercise that include closed kinetic chain activities, balance training, eccentric and high-repetition/low-load exercises, reflex facilitation through reactive training, stretch-shortening activities, and biofeedback training. These techniques produce adaptations in sensitivity of peripheral receptors and facilitate afferent pathways, agonist/antagonist coactivation, muscle stiffness, reflex muscle activation, and discriminatory activation of muscles. Alternatively, the superior balance and sense of joint position of gymnasts could be genetically determined.
The potential existence of genetic predisposition versus the effect of training has to be clarified. Although definite conclusions cannot be drawn from our study or previous studies, we agree with the postulate put forth by Barrack 2 and Lephart 19 : The effects of training on muscles and tendons may be the main factor for this enhanced sense of joint position and balance in gymnasts, because muscle receptors provide reliable proprioceptive information.
The clinical implications from this study are threefold. First, this study showed that the testing method was reliable, establishing a protocol for future proprioceptive testing and rehabilitation. Second, the results of this study suggest that extensive training has a positive influence on sense of position of the ankle-joint and on balance, in addition to increasing muscle tone. Gymnasts are able to develop enhanced neurosensory pathways, neuromuscular control and functional stability as a result of long-term athletic training. Lastly, limb dominance was shown to have no effect on proprioceptive ability and on balance in volunteers or in gymnasts. This finding establishes validity for using the uninvolved limb as a control or baseline for testing and rehabilitation to aid in establishing discharge criteria.
CONCLUSION
The superior sense of position of a joint and balance play an important preventive role in gymnasts who are at risk for recurrent sprains. Gymnasts in this study demonstrated a higher incidence of balance and superior sense of joint position in their ankle than did the nongymnasts as measured by their performance of the one-legged-standing balance task, the single-limb-hopping course and active and passive joint-position. Such information, in conjunction with clinical data, can provide clinicians with additional clinical insight for classifying or categorizing impaired conditions of posture and/or movement with a focus on causal elements. This information might be useful for identifying gymnasts who are at risk for recurrent ankle isprains. In addition, it is recommended that rehabilitative programs for athletes include balance training and, more specifically for gymnasts, programs for rehabilitative balance that incorporate elements of their gymnastic routines.
Footnotes
ACKNOWLEDGMENT
We would like to thank John D. Hsu, M.D., for kind assistance in the preparation of this manuscript.