Mechanisms of agonist and antagonist activation in the knee of individuals with anterior cruciate ligament reconstruction: kinetic and eletromyographic study

Abstracts

PURPOSE: To assess and compare torque and electromyographic activity of the vastus lateralis and biceps femoris muscles upon knee extension and flexion in open kinetic chain. METHODS: Fifteen male subjects were distributed in two groups: Test Group (TG) (32.2 ± 7.1 years) composed by five subjects who had previously been submitted to anterior cruciate ligament arthroscopic reconstruction (patellar tendon); and Control Group (CG) (30.1 ± 10.7 years) composed by ten uninjured subjects. Data acquisition was performed using Cybex 6000 at 100°.s-1; 10 seconds of electromyography data were obtained using active differential electrodes (Delsys-Bagnoli 8) at a sample rate of 1000 Hz. Root Mean Square (RMS) values and temporal pattern of muscles activation based on movement phase were considered (linear envelope). RESULTS: Injured legs showed greater flexor peak torque and smaller extension peak torque; greater agonist activation and smaller antagonist activation for the biceps femoris muscle, and smaller agonist activation for the vastus lateralis muscle. Linear envelope showed that test group showed smaller vastus lateralis muscle activation comparing to control group. CONCLUSION: Despite the rehabilitation period, injured legs still showed extensor torque deficit, which may explain the remaining complains presented by anterior cruciate ligament reconstructed subjects.

Anterior Cruciate Ligament; Physical therapy; Electromyography; Biomechanics; Knee

OBJETIVO: Avaliar e comparar o torque e a atividade eletromiográfica dos músculos vasto lateral e bíceps femoral durante a extensão e a flexão do joelho em cadeia cinética aberta. MÉTODO: 15 sujeitos do sexo masculino, distribuídos em: cinco no Grupo Teste (GT) (32,2 ± 7,1 anos) com reconstrução do ligamento cruzado anterior via artroscópica (tendão patelar), e dez no Grupo Controle (GC) sem lesão (30,1 ± 10,7 anos). Foi utilizado o Cybex 6000 a 100°.s-1 e eletrodos bipolares diferenciais ativos (Delsys-Bagnoli 8), com a freqüência de amostragem de 1000 Hz e tempo de aquisição de 10 segundos. Foram considerados os valores do Root Mean Square (RMS) e o padrão temporal de ativação dos músculos em função da fase do movimento (envoltório linear). RESULTADOS: O lado lesado apresentou maior pico de torque flexor e menor pico de torque extensor. Maior ativação agonista e menor ativação antagonista para o bíceps femoral e menor ativação agonista para o vasto lateral. Pelo envoltório linear a ativação do vasto lateral no grupo teste foi diminuindo. CONCLUSÃO: Apesar de reabilitados, o membro lesado permaneceu com déficits no torque extensor, apresentando menor, mais precoce e decrescente ativação do músculo vasto lateral e menor ativação antagonista do músculo bíceps femoral, apesar do maior torque flexor e da maior ativação de unidades motoras durante a flexão do joelho. Estes déficits podem explicar algumas queixas clínicas que permaneceram nestes indivíduos.

Ligamento cruzado anterior; Fisioterapia; Eletromiografia; Biomecânica; Joelho

UPDATE ARTICLE

Mechanisms of agonist and antagonist activation in the knee of individuals with reconstructed anterior cruciate ligament: a kinetic and eletromyographic study

Anice de Campos Pássaro^I; Amélia Pasqual Marques^II; Isabel de Camargo Neves Sacco^III; Alberto Carlos Amadio^IV; Tatiana de Almeida Bacarin^V

^IPhysical Therapist, Master of the Department of Physical Therapy, Speech Pathology and Audiology, and Occupational Therapy, University of São Paulo Medical School, Professor at the Centro Universitário Capital

^IIFull Professor, Department of Physical Therapy, Speech Pathology and Audiology, and Occupational Therapy, University of São Paulo Medical School

IIIPh.D. Professor, Human Motion and Stance Biomechanics Laboratory, Physical Therapy, Speech Pathology and Audiology, and Occupational Therapy, University of São Paulo Medical School

^IVChairman of the Biomechanics Laboratory, University of São Paulo Gymnastics and Sports School

^VPhysical Therapist, Master of the Human Motion and Stance Biomechanics Laboratory, Physical Therapy, Speech Pathology and Audiology, and Occupational Therapy, University of São Paulo Medical School

Correspondences to

SUMMARY

Purpose: To assess and compare torque and electromyographic activity of the vastus lateralis and biceps femoris muscles upon knee extension and flexion in open kinetic chain. Methods: Fifteen male subjects were distributed in two groups: Test Group (TG) (32.2 ± 7.1 years) composed by five subjects who had previously been submitted to anterior cruciate ligament arthroscopic reconstruction (patellar tendon); and Control Group (CG) (30.1 ± 10.7 years) composed by ten uninjured subjects. Data acquisition was performed using Cybex 6000 at 100°.s-1; 10 seconds of electromyography data were obtained using active differential electrodes (Delsys-Bagnoli 8) at a sample rate of 1000 Hz. Root Mean Square (RMS) values and temporal pattern of muscles activation based on movement phase were considered (linear envelope). Results: Injured legs showed greater flexion peak torque and smaller extension peak torque; greater agonist activation and smaller antagonist activation for the biceps femoris muscle, and smaller agonist activation for the vastus lateralis muscle. Linear envelope showed that test group showed smaller vastus lateralis muscle activation comparing to control group. Conclusion: Despite the rehabilitation period, injured legs still showed extensor torque deficit, which may explain the remaining complaints presented by anterior cruciate ligament reconstructed subjects.

Keywords: Anterior Cruciate Ligament; Physical therapy; Electromyography; Biomechanics; Knee.

INTRODUCTION

The anterior cruciate ligament (ACL) stabilizes tibial anterior displacement and the varus and valgus movements of this joint⁽¹⁾. In vivo studies show that direct load tensions on this ligament would produce inhibition of the femoral quadriceps favoring knee flexors. These findings may have a clinical significance, because knee functions should potentially be impaired by ruptures of those ligament mechanoreceptors^(2-5). Osternig et al.⁽⁶⁾ investigated the coactivation of quadriceps and ischiotibial muscles on the isokinetic dynamometer in speed racer and long-runner women, and found that knee flexors are more active during knee extension than during quadriceps flexion. This could be associated to a stronger quadriceps force, which would require a higher antagonist coactivation of flexors for coordinating and decelerating the limb. Additionally, the larger muscular mass of the quadriceps could also favor flexion deceleration by a viscoelastic effect, meaning that this flexion would be primarily controlled by a passive mechanism⁽⁶⁾. Nielsen and Kagamihara^(7,8) suggested the existence of a specific motor program for muscle coactivation in which the reciprocal model's interneurons would be actively inhibited by central commands. That is, there would be a drop of the reciprocal inhibition, which would increase the level of excitement of motorneurons of the antagonist muscles, resulting in coactivation. The same theory is also advocated by other authors^(9-12). Carolan and Cafarelli⁽¹⁰⁾ found that endurance drills on knee extensors resulted in less coactivation between extensors and flexors. Thereby, they believe that after a week of training, the reduction of the strength opposing to movement, would contribute to a significant increase of the voluntary maximum extension contraction. Literature shows evidences that the muscle coactivation occurring in the knee would act protecting and stabilizing the joint during powerful contractions, and also could distribute the pressure through the joint and minimize joint fatigue and damages^{(4-6,9,12-17)}.

Osternig et al.^.(4) found that the flexor muscles of ACL injured knees generated less antagonist activity when compared to the contralateral limb, and didn't find significant differences for muscle torque both in the group with ACL injury and in the control group. A recent study confirmed this theory⁽¹⁸⁾. Makihara et al.⁽¹⁹⁾ found lower flexor isometric and isokinetic torque values from 60° of knee flexion in individuals submitted to ACL reconstruction using the tendon of semitendinous and gracilis muscles. This study aimed to assess and compare the torque and electromyographic activity of vastus lateralis and biceps femoralis muscles between injured and uninjured members of individuals with reconstructed ACL to a group of healthy individuals at extending and flexing the knee in an open chain, thus contributing to rehabilitation protocols, potentially focusing more specific muscular work out, stimulating the mechanism of muscle coactivation.

MATERIALS AND METHODS

Subjects

Our sample was constituted of 15 male volunteer subjects in the age group of 20 to 40 years old, five with reconstructed ACL assigned to the Test Group (TG) and ten subjects showing no osteomyoarticular injuries or dysfunctions assigned to Control Group (CG). In order to compose the TG, we selected subjects submitted to ACL reconstruction through arthroscopy using patellar tendon at least eight months previously to data gathering, and they should have been clinically and physiotherapeutically discharged. All subjects signed a Free and Informed Consent Term as approved by the local Committee of Ethics (Research Protocol # 559/01).

Material

For torque assessments, the isokinetic dynamometer Cybex 6000 (New York, USA ) was used, and, for electromyographic (EMG) studies, active differential bipolar electrodes (Delsys-Bagnoli 8, Boston, USA) made of Ag/AgCl, with size of 10 X 1 mm, inter-electrodes distance of 10 mm and pre-amplification of 10 on the electrode and 100 on the conditioner, totaling a gain of 1000, were employed. The sampling frequency was 1000 Hz and the acquisition time was 10 seconds. A 12-Bit A/D converser was used. For identifying the motor point of the vastus lateralis and biceps femoralis muscles, a Universal Pulse Generator (Omni Pulse 900, Piracicaba, Brazil) generating 1-ms pulse trains in a tetanizing frequency (20 - 80 Hz) was used.

Evaluation protocol

From each patient's medical file we retrieved diagnostic, injury and surgery dates, injured knee, dominant knee and surgery type data. At interview, the most common complaint, time and place of rehabilitation were identified, proceeding to physical examination for inspection, palpation and evaluation of knee alignment. The motor points of vastus lateralis and biceps femoralis muscles were bilaterally identified by electric stimulation for subsequent placement of surface electrodes. These were fixed with double-face and transpore-type tape to preclude these electrodes to move during the exercise. The protocol employed for the Cybex was the execution of knee extensions and flexions at 100°.s-1, with a total range of motion of approximately 120° (0° corresponding to total extension). The subjects were positioned at sedestation with hips flexed at approximately 90°, and the rotational axis of the dynamometer was aligned to knee axis. Before acquiring the data, we established a familiarization step, in which the subjects would execute three knee flexion and extension movements followed by a rest period of 180 seconds previously to the assessment in order to avoid muscle fatigue. The subjects executed five repetitions at the selected speed, with the CG starting the test with the dominant limb, while the TG initiated it with the uninjured limb.

Data analysis

On the EMG, Root Mean Square (RMS) values obtained by means of sign rectification and by the selection of phases corresponding to knee flexion and extension for each muscle, and the temporal pattern of activation of the muscles as a function of the movement phase represented by the linear envelope were considered as a way to qualitatively represent the temporal coordination of muscle activity during movement. For obtaining linear envelopes, mathematic stages in a mathematic programming environment (Matlab) were developed using the following procedure: after removing the offset of the gross sign, the electromyographic signal was rectified by full wave, filtering was provided with a 4^th order low-pass butterworth-type filter with cut frequency of 5 Hz; then, the signal was normalized in intensity by its average and as a function of knee flexion and extension time (0 - 100% of the cycle). The following variables were extracted: 1) first activation peak of the vastus lateralis muscle; 2) end of the vastus lateralis muscle activation; 3) first activation peak of the biceps femoralis muscle; 4) end of the biceps femoralis muscle activation; 5) start of biceps femoralis muscle activation. (Figure 1).

Statistical analysis

The Shapiro Wilk's W-test tested the normality of distribution of all continuous variables in the present study. The variables were described for each experimental group in terms of averages and standard deviation. Comparative inter-group analyses were made comparing dominant limbs on the CG to the injured and uninjured limbs of the TG, and the non-dominant limb of the CG to the uninjured limb of the TG. In the intra-group analysis, we compared the dominant to the non-dominant limb of the CG, and the injured to the uninjured limb of the TG. The two-way ANOVA was used, with one factor being the experimental group and the other, the contralateral limb (repeated measurements), followed by Tukey's HSD test, being considered as significant the differences with a significance level of 5%.

RESULTS

The studied groups were found to present statistically similar results for age (CG=30.1±10.7 y.o., and TG=32.2±7.1 y.o.) and body mass index (BMI) (CG=23.6±2.1 kg, and TG=25.2±1.8 kg {(Student's t-test)}. The clinical characteristics of the TG are described on Table 1. All subjects were operated by the same surgical team, using the interference screw for fixating the graft. Also, they were rehabilitated at the same institution following a pre-established and well-recognized protocol.

Torque evaluation

In the inter-group analysis (Table 2), the flexor torque peak was higher for the injured side of the individuals in the TG when compared to CG subjects' dominant sides (p=0.049). Similarly, the flexor torque peak for TG subjects' uninjured side was higher when compared to CG subjects' non-dominant limb (p=0.017). In the intra-group analysis, for TG, the injured limb showed lower extensor torque peaks (p= 0.032), while for CG, no differences were found between limbs (Table 3).

Electromyographic evaluation

The inter-groups comparison of the RMS values can be found on Table 4. The TG showed a lower RMS of the biceps femoralis on the uninjured limb during its agonist phase compared to CG subjects' non-dominant limb (p= 0.016). When assessing the interaction effect between groups and sides, we found a statistically significant difference for biceps femoralis muscle activation both in the agonist (p=0.001) and antagonist phases (p=0.001). For vastus lateralis muscle, differences were found only for the agonist phase (p=0.001), thus being similar for the antagonist activation of this muscle (p=0.656). Therefore, in the intra-group analysis (Table 5), the CG showed a stronger agonist activation for biceps femoralis muscle on the dominant limb (p=0.024), while the TG showed statistically significant differences for both muscles, and a higher agonist activation (p=0.001) and a lower antagonist activation (p=0.001) of the biceps femoralis muscle on the injured limb and lower agonist activation of the vastus lateralis muscle on the injured limb (p=0.001).

Analysis of the Linear Envelope

Figures 2 and 3 illustrate the mean envelopes of the TG and CG for the muscles of both studied sides. We can see that the vastus lateralis muscle activation is reduced along its agonist action (0-50% on movement cycle) for the TG, especially on injured limb (Figure 2). In the same environment, CG shows a stable activation (Figure 3).

Table 6 shows the intra-group linear envelope of biceps femoralis and vastus lateralis muscles during agonist and antagonist phases. No statistically significant differences were found between CG subjects' limbs. In the CG, the injured limb was found to present the first activation peak of the vastus lateralis earlier than the uninjured limb (p=0.001). In the inter-group comparison, for linear envelope variables, the groups were shown to be statistically similar concerning muscle recruitment time pattern, except for the first vastus lateralis peak, which occurred earlier on the injured limb when compared to the dominant limb (p=0.001).

DISCUSSION

The theory behind this study was that individuals rehabilitated from ACL reconstruction did not show extensor torque differences, but indeed, differences concerning muscle activation pattern and intensity, especially associated to biceps femoralis muscle during its antagonist phase, as reported by literature^(4,6-10). However, the injured limb showed lower extensor torque peaks when compared to uninjured limbs, inconsistently with other studies^(4,18), but showing a recovery deficit in those individuals. Furthermore, in the inter-group analysis, we could also see a higher flexor torque peak on the injured limb compared to the dominant limb. Literature^(4,18) do not mention differences concerning flexor torque as well, and this fact is supposedly associated to rehabilitation programs focusing knee flexors strength gain, avoiding tibial anterior displacement forces, that could be tensioning the graft⁽²⁰⁾. Regarding muscle activation strength, we found a statistically significant difference between groups and sides for biceps femoralis muscle during agonist and antagonist phase. The injured limb showed higher agonist activation in this muscle, i.e., it recruits more motor units in this specific task, and it could be correlated to muscular training performed during rehabilitation. Additionally, only one flexor muscle was assessed by electromyography, with the joint torque being constituted by the action of several flexor muscles of the knee. Another important result, which is consistent with literature, is the lower antagonist activation of the biceps femoralis muscle in individuals with reconstructed ACL^(4-6,9-17). Even considering the limitations previously described, it is interesting to note that, despite of the higher flexion torque and of the higher agonist activation in this group of muscles - suggesting a well trained musculature in a joint stabilization moment - we can infer that this muscle structure was not properly activated. However, this can be associated to the lower extensor torque on the injured limb, as well as to the lower agonist activation of this muscle, meaning that this extensor moment was not enough to activate the antagonist muscles, and this extension deceleration may have been provided by the viscoelastic effect of flexor muscles of the knee⁽⁶⁾ . The lower antagonist activation of the biceps femoralis muscle in individuals with reconstructed ACL, as identified by the lower RMS values, is consistent with literature, such as on the reports by Osternig et al.⁽⁴⁾ and Solomonow et al.⁽⁵⁾ who noticed an important coactivation reduction in individuals with reconstructed ACL, and they justify this fact by the absence of mechanoreceptors in the graft, which would be responsible for activating knee flexors, regulating the extensor torque. Thus, we can infer, as did those authors, that such activation reduction may be a muscle control deficit in those individuals. Amiridis et al.⁽²¹⁾ justified the lower coactivation of the knee flexors in high-performance athletes when compared to sedentary individuals as a function of the training demands for muscle hypertrophy. Therefore, they infer that such reduced antagonist activation of the biceps femoralis muscle can also evidence a better prepared musculature, as oppositely to a deficit. Baratta et al.⁽¹³⁾ disagree with the fact that ischiotibial strength drills are responsible for the lower coactivation of this muscle. Conversely, they believe that a routine training program for this muscle in athletes would lead this population to present coactivation levels similar to healthy individuals.

This coactivation increase is believed to be a potential early response due to the increased muscular recruitment for strength gain, but not necessarily one could say that this coactivation increase will remain. The qualitative analysis of linear envelopes suggest a reduced activation of the vastus lateralis muscle on TG during knee extension movement, which, to a certain extent, was expected as a way to avoid shearing forces to the graft. It is interesting to notice that this pattern is reflected on the uninjured limb; The present study also found that the CG and the TG were shown to be similar to each other in terms of muscle recruitment time pattern as observed by linear envelopes variables, except for the first peak of the vastus lateralis muscle, which occurred earlier on the injured limb compared to CG individuals' dominant limb and TG subjects' uninjured limb. This earlier activation is likely to occur on the injured limb as a way to offset a progressive activation reduction that occurs throughout the extension movement (0-50% of the movement cycle). Some limitations were seen over the development of this study, including: a small sample in the TG group; the fact that the individuals were operated by different surgeons and rehabilitated by different physical therapists; however, it is worthy to highlight that the purpose of the study was to assess individuals who had already been discharged from clinical and physiotherapeutic programs showing the same status as the majority of the patients who, after an injury, receive surgical treatment, some physical therapy sessions, and are discharged. At this moment, the purpose was not to check if the surgical technique, the rehabilitation protocol, as well as the recovery of the donor area may have influenced the assessed variables. These variables may be considered in future studies.

CONCLUSION

Although rehabilitated, the surgically repaired limb remained with extensor torque deficits, showing a lower and decreasing activation of the vastus lateralis muscle and a lower muscle coactivation. These deficits may explain some of the clinical complaints that patients frequently remain presenting, such as quadriceps muscle atrophy, instability and pain. Possibly, an earlier strength work on the quadriceps muscle might be more significantly beneficial for ACL repair recovery by the improvement of the muscle coactivation mechanism.

REFERENCES

References

• 1. Amatuzzi MM. Antigos conceitos são modernos no tratamento das doenças ligamentares do joelho. Rev Bras Ortop. 2001; 36:1-8.
• 2. Barrack RL, Skinner HB, Buckley SL. Proprioception in the anterior cruciate ligament dificient knee. Am J Sports Med. 1989; 17:1-6.
• 3. Krauspe R, Schmidt M, Schable H. Sensory innervation of the anterior cruciate ligament. J Bone Joint Surg Am. 1992; 74:390-7.
• 4. Osternig LR, Caster BL, James CR. Contralateral hamstring (biceps femoris) coativation patterns and anterior cruciate ligament dysfunction. Med Sci Sports Exerc. 1995; 27:805-8.
• 5. Solomonow M, Baratta R, Zghou BH. The synergistic action of the anterior cruciate ligament and thigh muscles in maintaining joint stability. Am J Sports Med. 1987; 15:207-13.
• 6. Osternig LR, Hamill J, Lander JE, Robertson R. Coactivation of sprinter and distance runner muscles in isokinetic exercise. Med Sci Sports Exerc. 1986; 18:431-5.
• 7. Nielsen J, Kagamihara Y. The regulation of disynaptic reciprocal Ia inhibition during co-contraction of antagonistic muscles in man. J Physiol. 1992; 456:373-91.
• 8. Nielsen J, Kagamihara Y. The regulation of presynaptic inhibition during co-contraction of antagonistic muscles in man. J Pysiol. 1993; 464:575-93.
• 9. Snow CJ, Cooper J, Quanbury AO, Anderson JE. Antagonist cocontraction of the knee extensors during constant velocity muscle shortening and lengthening. J Eletromyogr Kinesiol. 1995; 5:185-92.
• 10. Carolan B, Cafarelli E. Adaptations in co-activation in response to isometric training. J Appl Pysiol. 1992; 73:911-7.
• 11. Psek JA, Cafarelli E. Behaviour of coative muscles during fatigue. J Appl Physiol. 1993; 74:170-3.
• 12. Snow CJ, Cooper J, Quanbury AO, Anderson JE. Antagonist cocontraction of the knee flexors during controlled muscle shortening and lengthening. J Eletromyogr Kinesiol. 1993; 3:78-86.
• 13. Baratta R, Solomonow M, Zhou BH, Letson D, D'Ambrosia R. Muscular coativation. The role of antagonist musculature in maintaining knee joint stability. Am J Sports Med. 1988; 16:113-22.
• 14. Draganich LF, Valey JW. Coativation of the hamstrings and the quadriceps during extension of the knee. J Bone Joint Surg Am. 1989; 71:1075-81.
• 15. Lee JB, Matsumoto T, Othman T, Yamauchi M, Taimura A, Kaneda E et al. Coactivation of the flexor muscle as a synergist with the extensors during ballistic finger extension moviment in trained kendo and karate athletes. Int J Sports Med. 1999; 20:7-11.
• 16. Solomonow M, Baratta R, Zghou BH, D'Ambrosia R. Electromyogram coativation patterns of the elbow antagonist muscles during slow isokinetic movement. Exp Neurol. 1988; 100:470-7.
• 17. Yanagawa T, Shelburne K, Serpas F, Pandy M. Effect of hamstrings muscle action on stability of the ACL-deficient knee in isokinetic extension exercise. Clin Biomech. 2002; 17:705-12.
• 18. Urabe Y, Kobayashi R, Sumida S, Tanaka K, Yoshida N, Nishiwaki GA et al. Eletromyographic analysis of the knee during jump landing in male and female athetes. Knee. 2005 ; 12:129-34.
• 19. Makihara Y, Nishino A, Fukubayashi T, Kanamori A. Decrease of knee flexion torque in patients with ACL reconstruction : combined analyses of the architecture and function of the knee flexor muscle. Knee Surg Sports Traumatol Arthrosc. 2006; 14 :310-17.
• 20. Chmielewski TL, Hurd WJ, Rudolph KS, Axe MJ, Snyder-mackler L. Perturbation training improves knee kinematics and reduces muscle co-contraction after complete unilateral anterior cruciate ligament rupture. physical therapy. 2005 ; 85: 740-50.
• 21. Amiridis IG, Martin A, Morlon B, Martin L, Cometti G, Pousson M et al. Co-activatin and tension-regulating phenomena during isokinetic knee extension in sedentary and highly skilled humans. Eur J Appl Physiol Occup Physiol. 1996; 73:149-56.

Publication Dates

• Publication in this collection
  04 July 2008
• Date of issue
  2008

History

• Accepted
  20 June 2007
• Received
  24 Apr 2007