Electromyographic activity of selected leg musculature in subjects with normal and chronically sprained ankles performing on a BAPS board.

GL Soderberg, PhD, PT, FAPTA, is Professor and Director, Physical Therapy Program, College of Medicine, The University of Iowa, 2600 Steindler Bldg, Iowa City, IA 52242 (USA). Address all correspondence to Dr Soderberg.

TM Cook, PhD, PT, is Assistant Professor, Physical Therapy Program, College of Medicine, The University of Iowa,

SC Rider, BS, and BL Stephenitch, MPT, PT, were students at The University of Iowa at the time the study was conducted.

This article is from a poster presentation at the joint Congress of the American Physical Therapy Association-Canadian Physiotherapy Association, june 12-16, 1988, Las Vegas, Nev.

This study was supported in part by a grant from Camp International.

The study protocol was approved by The University of Iowa's College of Medicine Committee on Human Subjects.

The Purpose of this study was to evaluate the magnitude and temporal features of tibialis anterior, peroneus longus, and gastrocnemius muscle activity during exercises performed on a Biomechanical Ankle Platform System (BAPSG) board by subjects with normal and chronically sprained ankles. Fourteen participants in each of the two groups rotated the platform in both clockwise and counterclockwise directions at self-selected velocities. Trials were completed with both medium- and large-sized hemispheres under the platform while electromyographic surface recordings were made from the three muscles of interest. Electromyographic values, normalized within individuals as a percentage of maximal voluntary isometric contraction (MVIC), ranged from 20% to 80% of MVIC An analysis of variance for the average myoelectric signals from all three muscles revealed no differences between the two groups Effects of direction and hemisphere size were shown for all muscles. The large hemisphere produced significantly greater average activity and a more pronounced effect on the tibialis anterior muscle. The data provide information on the level of muscle activity required for conditions used in this study and may assist the physical therapist with making judgments about the use of this device in rehabilitation programs, [Soderberg GL, Cook TM, Rider SC, Stephenitch BL. Electromyographic activity of selected leg musculature in subjects with normal and chronically sprained ankles performing on a BAPS(R) board. Phys Ther. 1991;7-1..514-522.]

Key Words: Electromyography, Lower extremity, Muscle performance.

Injuries to the supporting structures of the ankle joint are among the most common in the body. Garrick(1) has reported that 85% of these injuries are sprains involving the lateral structures.

Smith and Reischl(2) have reported that 70% of the athletes in their study had a history of an ankle sprain, with 80% of these athletes having multiple sprains.

The literature is laden with recommendations for strengthening ankle musculature as one means of preventing recurring sprains. Glick et al(3) and Karlsson et al(4) have proposed the use of a tilt board as part of their rehabilitation program. Exercises on a "sprain board"(5) and on balance boards for purposes of coordination and proprioceptive conditioning have also been proposed.(2) Gross(6) has suggested a model in which ankle joint receptors play a "dominant role in joint angle detection" and reported that judgments of joint positioning made when the muscles were relaxed were significantly better than judgments made when the muscles were contracted in a nonsprained group compared with subjects with previous ankle injury. Another study of subjects with inversion ankle sprains demonstrated that the response time of the peroneal musculature to sudden inversion stress was longer than, although not statistically different from, the response times of a noninjured control group.(7) Tropp et al(8) have indicated that athletes without previous histories of ankle injury or functional instability, as measured by stabilometry, demonstrated higher risks of sustaining ankle injuries than "normal players." Recently, ankle disk training has been shown to be beneficial for athletes with functional instability of the ankle.(9) Lentell et al(10) have shown that muscular weakness is not a major contributing factor to the occurrence of chronically unstable ankles.

No literature was identified that addresses the magnitude and temporal features of muscle activity while exercises are performed on boards designed to assist with rehabilitation and prevention of ankle injuries. This study was designed to evaluate the magnitude and temporal features of tibialis anterior, peroneus longus, and gastrocnemius muscle activity during exercises performed on the Biomechanical Ankle Platform System (BAPS(R)) board* by subjects with normal and chronically sprained ankles.

Method Subjects

Participants in this study were volunteers from the community, faculty, staff, and student population at The University of Iowa. All subjects signed informed consent forms. "Normal" subjects were those who reported no histories of ankle injuries. Thirteen women and one man comprised this group. The subjects' average height was 169.7 cm (66.8 in), and their average weight was 63.1 kg (138.9 lb).

The group of subjects with ankle injuries consisted of 4 women and 10 men. The subjects' average height was 182.9 cm (72.0 in), and their average weight was 82.3 kg (181.0 lb). Entrance into the study included an examination by one author (GLS) with experience as a physical therapist evaluating and treating sports injuries for 24 athletic teams for a period of 6 years. The mechanism of the original injury was described by the subject, as were the characteristics of the recurrent sprain. Descriptions and locations of tenderness were verified by the examiner, and all subjects in the recurrent- sprain group reported characteristics consistent with inversion ankle sprains. No special tests, such as drawer-sign tests, were used.

As a result of the evaluation, the examining therapist considered these subjects, whose recurrence of injury is described in Table 1, to collectively represent those individuals with chronic ankle sprain. Numerous subjects in the recurrent-sprain group were not informed of the classification of the sprain, and 8 of the 14 subjects lacked specific recall of the treatment rendered. All subjects had recovered sufficiently to participate in at least recreational sports. in all cases, the injured (right) lower extremity was tested.

Preparation

To record the subjects' myoelectric activity, the skin over the muscles of interest was wiped with alcohol-soaked gauze. Electrode assemblies were then attached using double-sided foam adhesive tape. Each assembly contained circuitry for preamplification with a gain of 35, measured 17 x 33 x 10 mm, and held two silver-silver chloride electrodes 8 mm in diameter. The distance between electrode centers was 22 mm. Conductive gel filled the holes in the tape over the electrode sites. The electrode for the tibialis anterior muscle was placed directly over the most prominent portion of the muscle belly just lateral to the anterior edge of the tibia. The electrode for the peroneus longus muscle was placed 7 cm directly distal to the head of the fibula. The gastrocnemius muscle's electrode was placed over the observable and palpable belly of the lateral head. In all instances, the long axis of the electrode assemblies was aligned with the direction of the muscle fibers. A common ground electrode was positioned over the right fibular head, and all electrode leads were plugged into main amplifiers.(+)

Instrumentation

The combined preamplifier and main amplifier system permitted a gain of 100 to 10,000 with a bandwidth of 40 Hz to 6 kHz. Common mode rejection ratio was 87 dB at 60 Hz, and input impedance was greater than 15 M[ohm] at 100 Hz. After amplification, the electromyographic (EMG) signals were processed to produce a root-mean-square (RMS) envelope with a time constant of 55 milliseconds. A standard BAPS(R) board, measuring 66 cm mediolaterally and 60 cm anteroposteriorly, was instrumented with two high-precision 1-k[ohm] potentiometers' fixed to the surface. Each potentiometer was powered by an HP 6234A power supply." The shafts were orthogonally located and attached such that the potentiometers measured board movement in the anteroposterior and mediolateral directions (Fig. 1). This output (in volts) was amplified and processed by a computer program." The combined output of the two potentiometers provided a unique signature that could be converted to depict the location of the board in space. Verification of the accuracy of the potentiometers for depicting the position of the unloaded, but rigid, board was completed before any testing. Angles of 0, 90, 180, and 270 degrees were marked on the board and verified by repeated measurements from the data- processing computer program. No statistical tests were applied to this data set, but completion of similar tests at the end of the data-collection sessions ensured the accuracy of the output.

Electromyographic signals from all channels were cabled to an IBM-PC computed' containing a high-speed analog-to-digital convertor.# Sampling was performed at a rate of 100 per second and coupled with the 55-millisecond time constant to avoid any possibility of undersampling of the EMG data. A representative output is shown in Figure 2. Although the EMG preamplifier system minimized undesired artifact, at least two channels of raw EMG data and both channels of potentiometer output were evaluated continuously for offsets and artifacts on a standard oscilloscope. All three channels of EMG data were evaluated periodically for each subject.

Rarely, a low-frequency offset was produced by contact of the opposite leg with the electrode over the peroneus longus muscle. In those instances, the trial was repeated.

Protocol

Because the EMG data were to be normalized, each participant's maximally evoked EMG signal for each of the specified muscles was recorded during isometric dorsiflexion, eversion, and plantar-flexion contractions requiring maximum effort from each of the muscles. Manual resistance was provided for dorsiflexion and eversion, but for plantar flexion the subject was positioned long-sitting between two fixed objects so that the ankle would be adequately resisted. For all contractions, the ankle and foot joints were maintained in a neutral position while the middle 3 seconds of the 5-second contraction was sampled. The participants were instructed to relax during the intervals of approximately 30 seconds while the data from each of three contractions for each of the three muscles were being stored on a computer disk. The contraction producing the greatest amount of EMG activity was used in determining the individually normalized data sets.

Each subject was then positioned on the BAPS(R) board for orientation to the exercises. The board was located inside a set of parallel bars to provide the necessary balance support. A mirror placed in front of the subjects minimized the necessity for them to flex the cervical spine to observe the wise CCW) direction. Preliminary testing established that 40 rpm was the most commonly used rate of board rotation. After demonstrating a constant board-rotation velocity of as near as possible to 40 rpm (usually after less than 1 minute of practice), the subjects rested for 2 minutes. Subjects then reassumed the position on the board and initiated rotations. A verbal reminder was given to minimize loading on the parallel bars through the hands. Once acceptably constant velocity of board rotation was reached, computerized data collection was initiated for an interval of 20 seconds, simultaneously sampling output from the potentiometers and three channels of EMG data. Although initiation of sampling relative to board position varied for each trial, we do not believe that this factor was significant, because board rotation was rather constant and because all data were analyzed according to the appropriate 10-degree interval for the number of board revolutions completed by each subject. The protocol was repeated for the other direction of board rotation. Trials in both directions were also completed for the other size of hemisphere. Selected subjects repeated the series of trials so that a reliability analysis could be completed.

Design and Data Analysis

The experimental research design was a three-factor, mixed design based on independent groups (ie, healthy subjects versus patients with chronic ankle sprains) and repeated measures according to type of exercise (ie, medium versus large hemisphere and CW versus CCW direction). Because the location of the board was known, the level of muscle activity could be determined and related to board position. All EMG values were normalized as a percentage of each subject's maximal voluntary isometric contraction (MVIC) for each muscle. The 360-degree rotation cycle was divided into 36 segments of 10 degrees.

Each subject's mean EMG output for all trials for each of the four combinations was calculated for each 10-degree interval. Data for each muscle were plotted, and descriptive statistics were calculated for each 10-degree interval. Coefficients of variation (CVs) were also determined for each muscle. An analysis of variance (ANOVA) on the average EMG signal from all 36 segments (10-degree intervalsx36=360[deg] total) was used to compare data across groups, hemisphere size, and direction of rotation and to test for interactions for each of the muscles.(11) A post hoc Bonferroni adjusted t test was used to compare within-muscle differences in average EMG signal among combinations of direction of rotation and hemisphere size.

Results

The 14 subjects with normal ankles completed an average of 14.6 (range =7.5-24.0, SD=4.3) board revolutions in the 20-second sampling period. The subjects with chronic ankle sprains completed 12.5 range=7.0-23.0, SD=3.4) revolutions. A two-sample equal-variance t test revealed no significant difference between the mean numbers of revolutions completed in 20 seconds.

The descriptive results for each muscle for hemisphere size and direction of rotation, plotted as the normalized EMG output (in percentage of MVIC) versus the 36 10-second segments of board position, are shown in Figures 3 through 5. Note that the time sequence was from left to right during CW rotation (top part of each figure) and from right to left during CCW rotation (bottom part of each figure).

The standard-deviation bars and the average CVs are indicative of the level of variability during the performance of the exercises. The ANOVA results for the tibialis anterior muscle's average EMG values are presented in Table 2. All two-way and three-way interactions were nonsignificant. Only the interaction of groupXdirection of rotation approached significance. Main effects were present for direction of rotation, hemisphere size, and subjects, but not for group. For the peroneus longus and gastrocnemius muscles, subjects were the only source that demonstrated statistical significance. Data for the average EMG values calculated across each condition are presented in Table 3. For the tibialis anterior muscle, three of the six possible comparisons were significant when submitted to a post hoc Bonferroni adjusted t test. For the peroneus longus and gastrocnemius muscles, none of the comparisons were statistically significant.

Discussion

The results of this study represent the most comprehensive analysis presented to date on the use of a balance board as a rehabilitative technique. No literature was located that allowed us to make comparisons with earlier work on muscles controlling the ankle under these test conditions. Considering the effort required to perform the exercises, the values reported appear reasonable. It is important to note that all EMG values were normalized as a percentage of MVIC. This, or a similar, normalization procedure is required to standardize EMG data across subjects. Because the contractions could be considered dynamic, allowing the muscle to both shorten and lengthen, caution should be used in making interpretations from these data.

As no literature is available that describes the response of healthy subjects or of patients with recurrent ankle sprains during balance-board rotation, the degree of variability of response to this activity is an important finding. When designing the study, we were aware of the potential for large amounts of variability in the responses we were eliciting. We elected, however, to allow self-selected velocity of board rotation, less- than-rigid control of pressure on the hands, and a short practice period to simulate the circumstances that clinicians would likely use with their patients.

Reflective of the high degree of variability are the standard deviations of the averaged EMG values for each interval, ranging from 13.8% to 43.7% of the mean values (Tab. 3). Also, as shown in Figures 3 through 5, the average CVs ranged from 49% for CW rotation for the tibialis anterior muscle to 97% for CCW rotation for the gastrocnemius muscle. Several explanations for the large degree of variability can be offered.

One explanation is that velocity of board rotation was not rigidly controlled. Most subjects completed between 12 and 18 rotation cycles during the 20-second data-sampling interval. We allowed self-selected velocity in order to simulate clinical conditions, and we believe that the practice trials were sufficient to ensure representative performances on the BAPS(R) board. That there were equal variances and no significant differences between the two groups minimizes any arguments that can be made as to the degree of importance of this variable. Velocity within a cycle of rotation, however, was free to vary.

Another factor influencing the standard deviations and CVs is the degree of support used by each subject. No specific mechanism for control of degree of support was used, but we exerted some control over the amount of support provided by the parallel bars by informally reminding the subjects to minimize the pressure being placed on the hands. Although recognizing the lack of specific control of this variable, we believe that the support was typical for a clinical setting and not a ma or influence on the results obtained for the use of the BAPS(R) board.

Motion of the whole body over the axis of the board could also be an influential factor. Other than verbal reminders to maintain consistent postures, subjects were able to exert minor posture shifts that may have assisted them with board rotation.

Considered collectively, the velocity of board rotation, degree of support, and joint position are acknowledged to affect the subjects' performance and resultant EMG values. Although not specifically controlled during the data- collection phases, the EMG data are believed to represent a true clinical result.

Although not a specific purpose of this study, we were also interested in assessing the reliability of subject performance. Ten subjects (3 with normal ankles and 7 with chronic ankle sprains) repeated trials within the same day. The Pearson Product-Moment Correlation Coefficient for between-replicate data pooled over all individuals and conditions showed that the level of .80 or higher was reached for 84% of the trials. Most of the .80 or higher coefficients resulted from the analysis for the tibialis anterior and peroneus longus muscles, rather than for the gastrocnemius muscle. This finding was as anticipated, because the CVs for the gastrocnemius muscle were 90% and 97% for CW and CCW rotation, respectively (Fig. 5). In examining the data, note was made of phase differences between the repeated trials (ie, the muscles were more or less active at different phases of board rotation). Although the high coefficients were probably attributable in part to the great dispersion in the values, the point remains that the balance-board activity generally appeared to be consistently performed on separate occasions.

Between-day correlations would likely have been lower. overall, trial-to-trial reproducibility indicated that a 20-second data-sampling interval was adequate to represent the level and duration of activity occurring in the muscles evaluated in this study.

The percentage of EMG activity from each muscle appears consistent with the magnitude of torques required on the balance board to produce rotation.

These torques are determined by location of the center of body mass relative to the axis of the board.

Although there are apparently no published data on these torques, our data were consistent with the requirements for board rotation. For example, the greatest amount of tibialis anterior muscle activity occurred when the board was most anteriorly inclined (ie, 330@-30*) (Fig. 3). This result is attributable to the fact that the ankle must begin dorsiflexing in order to elevate the anterior edge of the board. Similarly, the peroneus longus muscle was active toward the extreme of either medial or lateral board position. Note should be made, however, of the potential plantar-flexing effect of this muscle in the 10-degree intervals preceding the most lateral and medial board positions (Fig. 4). The position and magnitudes of these peaks are possibly a result of the degree of knee angle and the amount of varus and valgus produced by the subjects. No efforts were made to control for these factors because we wanted to determine the effect of hemisphere size and direction of rotation under normal conditions of performance. We were somewhat surprised that the gastrocnemius muscle activity was not greater when the board was in a posterior position (Fig. 5). Possible explanations are that other muscles were active or that a shift of body weight occurred during this phase of board rotation.

An ANOVA was completed for each muscle, because no across-muscle comparisons are considered legitimate with these EMG data. Because the ANOVA results for only the subjects variable were significant at the .05 level of probability, only the ANOVA results for the tibialis anterior muscle are presented in this article (Tab. 2). Although the groupXdirection interaction approached significance for the tibialis anterior muscle, the .05 level was not reached. We did not attach any special meaning to this lower probability, because no similar finding occurred for other muscles, nor could we find an explanation on examination of the data. Direction of rotation for the tibialis anterior muscle was highly significant (Tab. 3). As shown in Table 3, higher levels of tibialis anterior muscle activity appear to be required for CW rotation than for CCW rotation. The significant differences for direction, however, may result from different EMG phase relationships that depend on the direction of rotation (Fig. 3). It should be noted that the overall magnitudes of EMG activity were similar during these activities.

As may he anticipated, the size of the hemisphere also influenced the level of muscular activity, but, again, only for the tibialis anterior muscle. Explanations for the absence of other differences attributable to hemisphere size are elusive. We did observe, however, that subjects used slightly different angles of knee extension and that some subjects shifted the body more mediolaterally than others. More studies that control for these factors are needed to determine whether control of body posture alters the muscular requirements of the exercises our subjects completed on the BAPSO board. A comparison of values derived from exercises using the smaller hemisphere versus the larger hemisphere would also be of interest.

This study provides objective data on the phase and level of tibialis anterior, peroneus longus, and gastrocnemius muscle activity during exercise on a commercially available balance board used for rehabilitation of ankle and knee injuries. The study establishes that both the magnitude and duration of response do not differ when healthy subjects are compared with subjects with chronically sprained ankles.

Therapists should use this information to exercise patients in certain ranges of motion ROMS) of the balance board if they desire higher levels of activity in certain muscles. For example, if activity of the peroneus longus muscle is desired, therapists should consider exercising in selected ROMS, such as 150 to 240 degrees for CCW motion (Fig. 4). Because less muscular activity was required in all other phases, a small CCW and CW motion in this limited arc may produce the greatest effect on this muscle. Selective ROMs eliciting the greatest amounts of activity can be determined for the tibialis anterior and gastrocnemius muscles from Figures 3 and 5, respectively.

Physical therapy practitioners should note that hemisphere size has a more pronounced effect on the tibialis anterior muscle than on the peroneus longus or gastrocnemius muscle. Direction of rotation appears to have little effect on the magnitudes of EMG activity in the lower-leg musculature we studied, but therapists should be cognizant that the temporal sequence of the activity differs markedly depending on the direction of rotation.

Only further investigation of the effects of different hemispheres and the addition of weights to various locations on the BAPS(R) board will provide further information on applications of this device for purposes of rehabilitation of ankle injuries.

Conclusion

Subjects with normal and chronically sprained ankles responded similarly to the BAPS(R)-board exercises performed in this study. Electromyographic activity levels of the tibialis anterior, peroneus longus, and gastrocnemius muscles ranged from 20% to 80% of the effort evoked during MVICs. Although the larger hemisphere size had a greater effect than the smaller hemisphere size on the tibialis anterior muscle, direction of board rotation created temporal patterns that were different for each muscle.

References

1 Garrick JG. The frequency of injury, mechanism of injury, and epidcmiology of ankle sprains. Am J Sport. Med. 1977;5:241-242.

2 Smith RW, Reischl SF. Treatment of ankle sprains in young athletes. Am J Sports Med 1986;14:465 71.

3 Glick JM, Gordon RB, Nishimoto D. The prevention and treatment of ankle injuries. Am j Sports Med. 1976;4:13(o 141

4 Karlsson J, Bergsten T, Lansinger O, ct al. Reconstruction of the lateral ligaments of the ankle for chronic lateral instability. J Bone Joint Surg [Am]. 1988;70:581-588.

5 McCluskey GM, Blackburn TA, Lewis T. A treatment for ankle sprains. A J Sports Med 1976;4:158-161.

6 Gross MT, Effects of recurrent lateral ankle sprains on active and passive judgments of joint position. Phys Ther. 1987;67:1505-1509.

7 Nawoczcnski DA, Owen MG, Ecker Ml,, et al. Objective evaluation of peroneal response to sudden inversion stress. Journal of Orthopaedic a Sports Physical Therapy. 1985;7:107-109.

8 Tropp H, Ekstrand J, Gillquist J. Stabilometry, in functional instability of the ankle and its value in predicting injury. Med Sci Sports Exerc. 1984;16:64-66.

9 Tropp H, Askling C. Effects of ankle disc training on muscular strength and postural control. Clinical Biomechanics. 1988;3:88-91.

10 Lentell GL, Katzman LL, Walters MR. The relationship between muscle function and ankle stability. Journal of Orthopaedic and Sports Physical Therapy,. 1990;11: 605-611.

11 Snedecor GW, Cochran WG. Statistical Methods. 6th ed. Ames, Iowa: The Iowa State University Press; 1967.