Electromyographic biofeedback and recovery of quadriceps femoris muscle function following anterior cruciate ligament reconstruction.

A primary concern in the rehabilitation of postoperative joints is the recovery of range of motion. Anterior cruciate ligament (ACL) reconstruction is typically followed by a period of immobilization and inactivity that can result in atrophy of the knee musculature and adhesion formation. Both of these conditions can limit the ultimate recovery of knee ROM. Therefore, the immediate postoperative objectives are to regain ROM and to recover muscle control and strength. Regaining muscle control and performing muscle strengthening exercises, however, can be difficult during the early postoperative phase as a result of edema, pain, and the diminished activity of the joint receptors. [1] Feedback from these receptors influences the activation of motor units in the joint musculature. Therefore, any disruption of normal receptor feedback can interfere with the patient's ability to contract the surrounding musculature and consequently may impede the performance of rehabilitative exercise and the recovery of muscle control and strength and joint ROM.

The functional rehabilitation of the postoperative knee requires the recovery and balance of the quadriceps femoris and hamstring muscle groups. [2] Typically, the quadriceps femoris muscles suffer the greatest functional loss after ACL reconstruction and are the focus of rehabilitation protocols. [3,4] Two exercises commonly used by therapists to strengthen this muscle group are the quadriceps femoris muscle setting (QS) exercise and the straight-legraising (SLR) exercise. Both of these exercises are designed to facilitate quadriceps femoris muscle performance without imposing damaging stress on the graft site and suture line. As several clinicians have noted, however, patients often have considerable difficulty executing an effective QS exercise contraction following knee surgery, contracting primarily the hip musculature and neglecting to contract the knee extensors. [5,6] The patient often performs the SLR exercise incorrectly as well, using only the hip flexors to lift the leg without the additional contraction of the knee extensors to maintain knee extension.

Postoperatively, although pain and edema are certainly factors, the inability to perform these exercises correctly may also be the result of a temporary loss or distortion of proprioceptive feedback from the involved joint. This feedback provides information regarding muscle stretch, tendon tension, and joint position and is considered by motor control theorists to be a necessary component of motor skill acquisition. [7] Krebs et al have suggested that the tissue trauma associated with the anterior capsular incision used in surgical procedures involving the knee can cause a temporary distortion of anterior joint capsule receptor activity which may not return to normal for three weeks or more. [1] Because the anterior receptors facilitate quadriceps femoris muscle activity, the loss of these facilitory effects may result in diminished control and use of the quadriceps femoris muscles during rehabilitative exercises and, consequently, a limited rate of return to normal muscle function.

Several authors have suggested that the biofeedback maN, be a valuable augmentor of receptor feedback from the knee musculature during quadriceps femoris muscle exercises. [6,8-10]

This feedback signal serves as an immediate, precise, and concurrent source of information for the patient. [11] Therefore, although a patient may not be able to "feel" or perceive tile muscle activity he or she can see or hear the results of efforts to contract or relax the muscle. Feedback of this nature has been shown to play a crucial role in the acquisition of motor skills [12] and has facilitated significant clinical improvements in cases of postoperative hand rehabilitation, [13] shoulder joint instability, [14] spinal injury and stroke rehabilitation, [15-17] and postmeniscal repair. [18,19]

The purposes of the present study were to apply the principles of feedback-facintated muscle reeducation to specific quadriceps femoris muscle strengthening exercises in tile postoperative ACL protocol and to determine what effect the provision of more precise and correct feedback during exercise might have on the recovery of quadriceps femoris muscle function. Recovery of quadriceps femoris muscle function was assessed by 1) a dynamometric test of quadriceps femoris muscle isometric peak torque administered 12 weeks postoperatively and 2) the number of days that elapsed between surge and the recovery of full active knee extension. The hypothesis was that patients using biofeedback during quadriceps femoris muscle exercises would achieve a more rapid recovery of isometric peak torque and full active knee extension than patients who exercised without biofeedback.

Method

Subjects

Subjects for this study were 22 patients (15 male, 7 female) ranging in age from 16 to 36 years ([.sup.-]X = 23 years). All patients had acute injuries to the ACL that required an open surgical reconstruction using a bone-tendon-bone patellar tendon autograft procedure. During a postoperative checkup at one week postsurgery, patients were familiarized with the testing and training protocol. Those who agreed to participate in the study signed an informed consent statement and were then assigned to one of two groups: 1) a Control Group (n = 11), which followed the ACL rehabilitation protocol that was in use at Knoxville Orthopedic Clinic at the time this study was conducted (Tab. 1), or 2) a Treatment Group (n = 11) which followed the same rehabilitation protocol but also received biofeedback to monitor the performance of the QS and SLR exercises in the protocol. The patients were assigned to groups with a randomized matching procedure that controlled for age, gender, and nature of surgical repairs (ligament vs ligament and meniscus). All patients were fitted with a hinged brace to restrict ROM for the first six weeks. Brace settings, which were progressively less restrictive from Week 1 to Week 6, were similar for each group.

Instrumentation

The instruments used in this study included a Cyborg Model J33 portable biofeedback unit and a Cybex II isokinetic dynamometer with dual-channel chart recorder. The Cyborg Model J33 biofeedback unit has an integrated electromyographic signal and a bandpass width of 350 Hz and is calibrated to record muscle activity of up to 1,000 UV. it also has both a visual meter display and an audible signal that sounds when the patient exceeds a predetermined threshold level. The Cyborg Model J33 biofeedback unit was used by patients during the QS and SLR exercises as a monitor and facilitator of extensor use. Disposable surface sensors (Quick-Stik) were used. These sensors consisted of three electrodes embedded in a 10-cm-long adhesive strip with a 6-cm space between active electrodes. In preparation for the electrode placement, the skin just proximal to the patella was scrubbed with an alcohol pad. The electrode strip was placed just proximal to the patella and approximately 2 cm medially so as to focus on knee extensor activity and diminish the recording of hip flexor activity. The Cybexe II dynamometer was used to record the torque (ft-lb) produced by patients during the 12-week postoperative evaluation of quadriceps femoris muscle isometric peak torque. The chart recorder torque damping was set at 2. The torque channel was readjusted to the baseline value before each test at a different angle. Dynamometer torque calibration was performed on a regular monthly basis, but was not verified at each testing session. One tester (VD) was used for all patients. Torque values were not gravity-corrected.

Procedure

All patients initially met with the experimenter one week following ACL reconstruction. During this first postoperative therapy session, Control Group patients were reviewed on the proper performance of QS and SLR exercises, attempted the exercises, and were given verbal feedback as to the quality of their performance. An attempt was made to limit verbal cues to 1) "Press the back of your knee into the table and raise your heel off the table"; 2) "try to relax the hip muscles and concentrate on moving your knee and your kneecap"; and 3) during the SLR exercise, "maintain a QS exercise contraction all the way through the lifting and lowering of the leg. "

The Treatment Group followed a similar procedure, but was also introduced to the concept of biofeedback and instructed in the use of the biofeedback unit as a monitor of their exercise efforts. After electrodes were affixed and a baseline activity level determined, a threshold EMG value was set as an exercise goal. The initial selection of threshold value was different for each patient (because of individual differences in amount of subcutaneous tissue, edema, and ability to contract the quadriceps femoris muscle) and was set so that the patient had to contract primarily the knee extensors and exert maximally to reach the threshold. Patients were told that it would be more worthwhile to have the goal somewhat high and miss a few times than to have the goal so low that it was too easily attained. (Threshold values initially averaged 10-15 [micro]V, and they improved to 80-140 [micro]V over the 12-week follow-up as pain and edema subsided and muscle control increased.) During this first session, patients in both groups performed 10 QS exercises and 10 SLR exercises. The same format with increased repetitions was repeated during the Week 2 postoperative session; new baseline and threshold values were set for the Treatment Group.

When patients returned for the Week 4 postoperative appointment, a two-or three-day-a-week rehabilitation schedule was initiated. Throughout the 12-week program, both groups followed the same rehabilitation protocol (Tab. 1) (ie, between-group conditions were the same except for the absence or presence of biofeedback). All patients began each session with SLR exercises five sets of 10 repetitions each in the prone, supine, and side-lying positions). The Treatment Group performed this segment of the exercise routine while receiving biofeedback regarding the quality of their distal quadriceps femoris muscle contraction. Threshold settings were reevaluated at each session and reset if necessary. Control Group patients performed the SLR exercises without biofeedback, but were corrected verbally if, during the first few lifts, their technique appeared incorrect (ie, they were co-contracting quadriceps femoris and hamstring muscles or were exhibiting extensor lag).

Data Collection and Analysis

All patients participated in a total of 18 or 26 therapy sessions, depending on their weekly schedule. After the SLR exercise segment of each therapy session, active knee extension measurements were taken. For the gonio-metric measurement of active knee extension, the patients were seated on a table with their legs extended and with their hips at approximately 70 degrees of flexion. Functional progress was represented by the number of days that elapsed between the surgery date and the date that the patient was able to achieve full active knee extension.

During the 12th postoperative week, quadriceps femoris muscle isometric peak torque was determined using the Cybex II isokinetic dynamometer. The test involved three maximal 3-second contractions with 10-second rest periods between each of three angles (90, 60, and 45 degrees ). Testing of quadriceps femoris muscle strength was not performed at angles of less than 45 degrees because of strain imposed on the ACL graft at these angles. [20] Each patient was tested while in a seated position with the knee fixed at the test angle. The input shaft of the dynamometer was aligned with the knee joint axis, and the lever arm shin pad was positioned just above the malleoli. Patients were allowed to grasp the sides of the Cybex test table during the test. verbal commands of "ready" and "go" were given to the patients, and contraction time was counted aloud by the experimenter. No other verbal encouragement was given. Each patient's nonoperative limb was tested first. For each limb, the best of the three contractions was recorded. The ratios of peak torque in the operative to nonoperative limbs were recorded for each patient at each of the three angles. These data were analyzed using a 2 x 3 factorial analysis of variance (ANOVA) (groups x angles) for repeated measures on the second factor. The measure of quadriceps femoris muscle recovery, was determined to be the percentage of nonoperative limb torque attainable in the operative limb. [9,21]

Results

Results of the ANOVA revealed a significant main effect for groups (F = 8.11; df = 1,20;p < .01) (Tab. 2). A plot of the means for peak torque ratios shows that the Treatment Group achieved greater torque recovery throughout the tested ROM (Figure). None of the within-patients effects were significant, suggesting that peak torque remained fairly constant for patients in both groups throughout the tested ROM.

The rate at which patients regained ROM was determined by the number of days that elapsed between the surgery date and the date that the patient achieved full active extension. The mean number of days was computed for each group and compared using an independent-samples t test. The results shown in Table 3 revealed that patients in the Treatment Group achieved full active extension in a significantly shorter period of time than did patients in the Control Group (t[20] = 2.28,p = .033).

Discussion

The present study examined the effects of a regularly administered biofeedback protocol on the recovery of quadriceps femoris muscle function following ACL reconstruction. Quadriceps femoris muscle function was defined in terms of isometric peak torque and joint ROM (knee extension). The findings show a greater recovery of quadriceps femoris muscle function among patients who used biofeedback as a facilitator of QS and SLR exercises than among those patients who exercised without biofeedback.

Isometric Torque Ratios

After 12 weeks of rehabilitation, patients in the Treatment Group were able to generate a greater percentage of their "normal" (nonoperative limb) quadriceps femoris muscle peak torque in their operative limb than were patients in the Control Group at each of the three tested angles (90 degrees , 60 degrees, and 45 degrees ). These results concur with those of earlier work by Lucca and Recchiuti, who found significantly greater increases in quadriceps femoris muscle peak torque among isometrically trained subjects given biofeedback than among subjects who exercised without biofeedback. [10] Similarly, Krebs found greater increases in EMG levels associated with isometric tension among postmeniscectomy patients who performed QS and SLR exercises with the addition of biofeedback compared with patients who performed the exercises without biofeedback. [19]

Over 20 years ago, Hettinger argued that a greater work effort rather than the duration of the effort is the primary contributor to strength gains. [22] Going through the motions of an exercise does not guarantee that muscle strengthening will take place. The targeted muscle must be repeatedly overloaded, which often does not occur during rehabilitative exercise. Because of pain and the distortion of proprioceptive feedback, patients may he unaware that they are not using the proper muscle or that they are not exerting a maximal effort. Biofeedback can be used to increase a patient's awareness of muscle activity by providing additional information regarding the quality and magnitude of his or her contractions. This increased awareness could conceivably promote a greater work effort. If the additional information made available via biofeedback results in a more consistently focused work effort, greater torque recovery would he expected to occur among patients using biofeedback.

Range of Motion

Although the recovery of muscle strength is basic to functional performance, a more visible and indicative measure of functional recovery is the ability of the muscle to move the associated joint through the full ROM. The quadriceps femoris muscles function as extensors of the knee joint. The ability to actively and fully extend the knee is required for a correct and balanced gait pattern.

The patients in this study who received biofeedback during quadriceps femoris muscle exercises achieved full active extension significantly sooner ([.sup.-] X = 63 days) than those not receiving biofeedback ([.sup.-] X = 78 days) (p < .05). (In light of advances in surgical procedures and changes in postoperative protocols that have taken place since this project was initiated, this time frame may appear somewhat lengthy. Our protocol, however, included six weeks in a hinged knee brace with restricted extension.) Because active knee extension is dependent on the patient's ability to effectively contract the quadriceps femoris muscle and patients in the Treatment Group exhibited greater quadriceps femoris muscle peak torque recovery than patients in the Control Group, it is not surprising that these same patients were able to fully extend their knee sooner. These findings concur with those of Sprenger and colleagues, who reported significant increases in knee extension ROM as a result of using biofeedback during quadriceps femoris muscle exercises following a meniscectomy. [18]

Wolf and Binder-Macleod reported similar improvements in joint ROM of the upper and lower extremities of chronic stroke patient: who used biofeedback to facilitate muscle recruitment and relaxation. [16,17] Although the ROM improvements they observed were accompanied by increased EMG levels, these patients did not necessarily demonstrate a significant improvement in manual or locomotor task performance. It should be noted, however, that individuals who have suffered a stroke have neurological dysfunction along both affector and effector pathways. In such cases, augmented feedback from the periphery would be expected to have little impact on the central nervous system mechanisms governing coordinated limb movement. [17] In contrast, neuromuscular impairments that are primarily the result of peripheral receptor dysfunction, should show functional, task-related improvements with the addition of augmented and refined joint receptor information provided by biofeedback. The functional status of the joint is determined to a great extent by active ROM. In the present study, the rate of which active joint ROM was recovered was enhanced with the use of biofeedback.

Specificity of Treatment Effects

Rehabilitation research has consistently demonstrated the positive effects of using EMG biofeedback to facilitate muscle reeducation. [6,8-10,13-19] Whether these effects are a direct result of the learning that is inherent with any information-based training technique such as biofeedback or whether they are more closely related to the motivational factors that may result from the performance feedback made available by biofeedback, however, remains undetermined. As Wolf suggests, separating the specific effects of training from more general effects such as motivation requires a double-blind study design that would he extremely difficult to use when rehabilitating patients in a clinical environment. [23] Consequently, the impact of both learning and motivation on the results of this study must be considered.

Learning has been defined as a relatively permanent change in the individual's ability to respond and is the result of training, practice, and experience. For learning to take place, certain information regarding the correctness of a response is necessary. [24] When a patient is trying to relearn a motor response and regain voluntary control over impaired muscle, EMG biofeedback may supply the additional information needed when joint feedback is distorted and incomplete. Technically, the biofeedback unit monitors and signals the magnitude of motor unit recruitment during muscle contraction. For the patient, however, it is purely a means of confirming and reinforcing the fact that a correct motor response was made. Although the ultimate goal for the patients in the present study was to recover full use of their operative quadriceps femoris muscle (ie, peak torque generation and full active knee extension), the initial goals were simply to relearn to contract the quadriceps femoris muscle and then to use it properly during exercise. The feedback from the EMG monitor provided patients with immediate, qualitative information regarding the results of their efforts to contract the quadriceps femoris muscle during exercise and therefore facilitated a relearning process.

The literature examining the influences of feedback on motor task performance has established the positive motivational effects that knowledge of results has on the performer. [24] More recently, Hald and Bottjen demonstrated that subjects provided with visual feedback regarding their efforts on an isokinetic maximal contraction test produced significantly more torque than subjects who did not receive feedback. [8] Visual feedback from the isokinetic dynamometer provided subjects with a frame of reference within which to assess their performance. Performance assessment has been described as an important source of motivation. [25] When individuals are more keenly aware of their performance level, they are more impelled to set and strive for performance goals. Electromyographic biofeedback provides the same type of information regarding a patient's exercise performance and could be perceived as a motivating modality. When patients are provided a visual or auditory representation of otherwise covert or masked muscle activity and are given a quantitative goal and a means to monitor their efforts toward that goal, exercise effort and outcome are enhanced. in the present study, the patients who used biofeedback may have been more aware than the controls of the quality and magnitude of their muscle contractions during exercise and consequently were better able to set and achieve exercise goals. A higher and more consistent intensity of exercise would result in greater muscle recovery.

Limitation

A limitation of this study was the fact that the data collector was not blind to the treatment conditions. This control was not possible in this particular research setting.

Conclusions

The results of this study demonstrate that the regular use of EMG biofeedback during muscle strengthening exercises significantly improves the rate of functional recovery of the quadriceps femoris muscle following major ligament reconstruction. The use of biofeedback training has been easily incorporated into not only our present ACL rehabilitation protocol, but other postoperative knee treatment programs as well. When a patient exhibits signs of poor quadriceps femoris muscle control or when a patient does not appear to be contracting the muscle maximally, biofeedback training has proven to be a very effective tool. Although it may take some time for the therapist to teach the patient how to perform a specific exercise in conjunction with the biofeedback unit, the biofeedback device enables the patient to exercise and to monitor each repetition without assistance. Perhaps a more important indication for using biofeedback is the reaction of patients to this modality. Patients consistently communicate that using the feedback unit during exercise makes them better aware of whether they are exercising correctly and increases their feelings of muscle control during exercise.

Acknowledgments

I would like to thank Tommi Stubbs, PT; Lea Adcock, PT; Lori Ballard, ATC; and Chip Ladd, FT, ATC, for their help and cooperation during this project. Also, thanks go to Craig Wrisberg, PhD, doctoral committee head.

References

1 Krebs DE, Staples W, Cuttita D, et al: Knee joint angle: Its relationship to quadriceps femoris activity in normal and postarthrotomy limbs. Arch Phys Med Rehabil 64:441-447, 1983

2 Paulos L, Noyes FR, Grood E, et al: Knee rehabilitation after anterior cruciate ligament reconstruction and repair. Am J Sports Med 9:140-147, 1981

3 Sisk TD, Stralka SW, Deering MB, et al: Effect of electrical stimulation on quadriceps strength after reconstructive surgery of the anterior cruciate ligament. Am j Sports Med 15:215-219, 1987

4 Antich TJ, Brewster CE: Modification of quadriceps femoris muscle exercises during knee rehabilitation. Phys Ther 66:1246-1250, 1986

5 Noyes FR, Mangine RE, Barber S: Early knee motion after open and arthroscopic anterior ligament reconstruction. Am j Sports Med 15:149-160, 1987

6 Soderberg GL, Minor SD, Arnold K, et al: Electromyographic analysis of knee exercises in healthy subjects and in patients with knee pathologies. Phys Ther 67:1691-1696, 1987

7 Mulder T, Hulstyn W: Sensory feedback therapy and theoretical knowledge of motor control and learning. Am j Phys Med 63:14341438, 1984

8 Hald R, Bottjen E: Effect of visual feedback on maximal and submaximal isokinetic test measurements of normal quadriceps and hamstrings. Journal of Orthopaedic and Sports Physical Therapy 9:86-93, 1987

9 Delitto A, Rose SJ, McKowen JM, et al: Electrical stimulation versus voluntary exercise in strengthening thigh musculature after anterior cruciate ligament surgery. Phys Ther 68:660-663, 1988

10 Lucca JA, Recchiuti SJ: Effect of electromyographic biofeedback on an isometric strengthening program. Phys Ther 63:200-203, 1983

11 Basmajian JV: Introduction: Principles and background. in Basmajian JV (ed): Biofeedback: Principles and Practice for Clinicians. Baltimore, MD, Williams & Wilkins, 1983, pp 1-4

12 Mulder T, Hulstyn W, Van der Meer J: Electromyographic feedback and the restoration of motor control. Am J Phys Med 65:173-188, 1986

13 Brown DM, Nahai F: Biofeedback strategies of the occupational therapist in total hand rehabilitation. in Basmajian JV (ed): Biofeedback: Principles and Practice for Clinicians. Baltimore, MD, Williams & Wilkins, 1983, pp 90-106

14 Beall MS, Diefenbach G, Allen A: Electromyographic biofeedback in the treatment of voluntary posterior instability of the shoulder. Am J Sports Med 15:175-178, 1987

15 Nacht MB, Wolf SL, Coogler CE: Use of electromyographic biofeedback during the acute phase of spinal cord injury. Phys Ther 62:290-294, 1982

16 Wolf SL, Binder-Macleod SA: Electromyographic biofeedback applications to the hemiplegic patient: Changes in lower extremity neuromuscular and functional status. Phys Ther 63:1404-1413, 1983

17 Wolf SL, Binder-Macleod SA: Electromyographic biofeedback applications to the hemiplegic patient: Changes in upper extremity neuromuscular and functional status. Phys Ther 63:1393-1403, 1983

18 Sprenger CK, Carlson K, Wessman HC: Application of electromyographic biofeedback following medial meniscectomy: A clinical report. Phys Ther 59:167-169, 1979

19 Krebs DE: Clinical electromyographic feedback following meniscectomy: A multiple regression experimental analysis. Phys Ther 61:1017-1021, 1981

20 Renstrom P, Arms SW, Stanwyck TS, et al: Strain within the anterior cruciate ligament during hamstring and quadriceps activity. Am J Sports Med 14:83-87, 1986

21 Mendler HM: Knee extensor and flexor force following injury. Phys Ther 47:35-45, 1967

22 Hettinger T: Physiology of Strength. Philadelphia, PA, Charles C Thomas, Publisher, 1961, pp 19-20

23 Wolf SL: Electromyographic biofeedback applications to stroke patients: A critical review. Phys Ther 63:1448-1459, 1983

24 Salmoni A, Schmidt R, Walter C: Knowledge of results and motor learning: A review and critical reappraisal. Psychol Bull 95:355-385, 1984

25 Cratty BJ: Psychology in Contemporary Sport. Englewood Cliffs, Nj, Prentice-Hall, Inc, 1983, pp 48-67

(Tables and other figures omitted)