- Email: [email protected]
Peroneal reaction times for diagnosis of functional ankle instability
Foot and Ankle Surgery 2000
6: 31–38
Peroneal reaction times for diagnosis of functional ankle instability D. ROSENBAUM,∗ H.-P. BECKER,† H. GERNGRO߆ AND L. CLAES‡ ∗Motion Analysis Laboratory, Orthopaedic Department, University of Mu¨nster, Mu¨nster, †Surgical Department, Military Hospital, Ulm and ‡Department of Orthopaedic Research and Biomechanics, University of Ulm, Ulm, Germany
Summary In order to differentiate between functional and mechanical chronic ankle instability, peroneal muscle function was examined in sudden inversion movements produced by a 30° tilting platform. Surface electromyographic (EMG) signals were recorded from the peroneus brevis and longus, tibialis anterior, and soleus muscles. An electrogoniometer was used to record rearfoot motion. Sixty-five subjects were assigned either to a functional (n = 35) or a mechanical instability group (n = 30) based on radiographic and mechanical stability measurements. Mechanical instability was characterized by a greater talar tilt and anterior drawer in the unstable leg. The groups did not differ significantly in the stability characteristics of the unaffected leg. Functional instability was reflected in significantly longer reaction times of the peroneus brevis (65 versus 57 ms, P = 0.002) and longus muscle (60 versus 54 ms, P = 0.01) in comparison with the mechanical instability group. Therefore, in those unstable patients without a clear mechanical insufficiency, the problem appears to be caused by a neuromuscular instability. Peroneal reaction times may help to distinguish between different causes of chronic ankle instability. Keywords: chronic ankle instability; peroneal reaction time; surface electromyography; tilting platform
Introduction Despite adequate treatment of acute lateral ankle ligament injuries, chronic instability–mechanical or functional–occurs in ≈10–20% of the patients [1, 2]. Mechanical instability is caused by a ligament Correspondence: Priv. Doz. Dr Dieter Rosenbaum, Funktionsbereich Bewegungsanalytik, Klinik und Poliklinik fu¨r Allgemeine Orthopa¨die, Westfa¨lische Wilhelms-Universita¨t Mu¨nster, Albert-Schweitzer-Str. 33, D-48129 Mu¨nster, Germany (email: [email protected]). 2000 Blackwell Science Ltd
deficiency and refers objectively to clinical and radiographic stress tests. Functional instability comprises the patient’s subjective feeling of repeated giving way based on proprioceptive disorders or muscle weakness [2, 3]. However, it is often difficult to distinguish between these two entities because a mechanical instability may arise from functional instability. Recurrent inversion injury of the ankle may be related either to mechanical or functional factors alone or to both. Since they are inconsistently associated, it is clinically difficult to decide whether
31
32
D. ROSENBAUM ET AL.
surgery is needed or not, especially in cases with borderline radiographic findings. Presently, no objective measure exists for the diagnosis of functional instability and it is generally dependent on the physician’s judgement alone. Before deciding upon a surgical reconstruction in a patient with chronic instability, a rehabilitation program is implemented. Approximately 50% of the patients will regain satisfactory functional stability after 12 weeks [4]. The remaining 50% will lose valuable 3 months with the risk of new injuries. Therefore, it is mandatory to seek useful tests for the diagnosis of functional instability and proprioceptive disorders. Freeman et al.[5] were the first to report on functional ankle instability. They confirmed the diagnosis using the clinical Rhomberg-test and demonstrated that strengthening and co-ordination exercises were able to relieve the symptoms. In athletes, Tropp found muscle weakness and an impaired ability to maintain postural equilibrium to be associated with functional instability [3]. Recently, trapdoor measurements have been used to investigate the reaction time of the peroneal muscles to an inversion movement [3,6–14]. Some of these studies demonstrated a generally slower reaction time in unstable ankles [11, 14, 15]. There is no study, however, which uses the information on muscle function for the diagnosis of functional instability and subsequent clinical decision making. The aim of the present study was to evaluate whether reaction time of the peroneal muscles provoked by sudden inversion movement is a useful parameter to distinguish between mechanical and functional instability.
Materials and methods Sixty-five patients with self-reported lateral ankle instability volunteered to participate in this prospective study. Thirty-five patients had been excluded due to bilateral symptoms, hyper-mobility and severe pain problems. A detailed questionnaire was used to evaluate the patients’ history of instability and the frequency of inversion traumata. Clinical and radiographic examinations of the ankle joint complex were analysed by the same physician (H.P.B.). Stress radiographs measuring anterior drawer and talar tilt were performed using a Scheubaapparatus (Telos Co., Hungen, Germany).
Based on the results of the radiographic and mechanical stress analyses, the subjects were separated into one of the following groups: (1) MI = mechanical instability (n = 30); and (2) FI = functional instability (n = 35). Three of the following criteria had to be fulfilled for a subject to qualify for the MI group:
1. More than eight inversion traumata per month. 2. Joint instability in the manual drawer and inversion test clearly increased in comparison to the contralateral side. 3. Radiographic talar tilt >7° and the difference between unstable and contralateral ankle >0°; 4. Radiographic anterior drawer >7 mm and the difference between unstable and contralateral ankle >0 mm. In order to evaluate the diagnostic value of the applied procedure the ligament status of those mechanically unstable patients (n = 26) who agreed to surgical treatment was determined intraoperatively. The same surgeon inspected the lateral ligaments (ATFL and CFL) and described it either as stable (intact fibrous structures or strong scar tissue) or unstable (elongated or defect). The findings were compared with the results of the reaction time measurements. A tilting platform with a mechanical release mechanism was built (Figure 1). The platform was 20 cm high and had a 44×40 cm wooden surface. The tilting part was 19 cm wide and was mounted on roller bearings. An aluminium lath was aligned with the lateral edge of the foot in order to prevent slipping. The trapdoor was released by pushing a foot switch not visible for the subject. Upon release the platform fell down through an arch of 30° which was predetermined by a mechanical stop (maximum angle of 45°). A potentiometer aligned with the axis of rotation was used to record the onset and duration of the tilting movement. After preparation the subjects were asked to stand on the titling platform. Six dynamic inversion measurements were executed in a random order in both extremities. In order to simulate an inversion trauma more realistically the patients were asked to shift the body weight onto the tilting foot with the contralateral foot only slightly planted to keep balance, i.e. almost full body weight loading. 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
PERONEAL REACTION TIMES
33
Figure 1 Schematic of the experimental set-up with a rear view of the subject’s lower legs and the electrogoniometer fixed to the heel and calf in line with the longitudinal leg axis.
Electromyographic (EMG) signals were recorded from four lower leg muscles – i.e. the peroneus longus, peroneus brevis, tibialis anterior and soleus – by means of custom made surface electrodes with on-site preamplification (gain factor 10). The signals were fed into a battery operated amplifier with a gain factor of 100, an input impedance over 1010 Ohm, a frequency range from 15 to 1056 Hz, and a common mode rejection ratio of 84 db. The electrodes had two silver detection surfaces with a diameter of 12 mm and an interelectrode distance of 20 mm centre-to-centre. In order to reduce skin impedance the recording sites were prepared by shaving, cleaning with 70% alcohol, and applying electrode gel. The electrodes were fixed with double-adhesive washers on the skin over the respective muscles. The muscles were palpated during contraction. The peroneus longus electrode was located about 3 cm below the proximal head of the fibula, the peroneus brevis electrode about 5 cm above the lateral malleolus. The tibialis anterior electrode was fixed 1 cm lateral to the tibial edge, the soleus electrode right below the head of the gastrocnemius near the peroneus brevis electrode. After preparation the electrode fixation and signal quality were controlled for movement and cable artifacts by tapping against the electrode. Correct positioning was controlled by asking the subject to perform isolated foot movements in order to elicit activity from the 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
respective muscles – peroneus brevis and longus: eversion movements with the foot planted; tibialis anterior: lifting the forefoot; soleus: raising on the toes. This procedure also allowed for a qualitative control of crosstalk artifacts since during isolated muscle activity in one EMG channel the other channels should remain silent. Foot eversion/inversion measurements were recorded with a strain-gauge goniometer (Type M 110, Penny & Giles, Blackwood, UK). The distal end block was taped to the calcaneus and the proximal end block was taped to the calf in line with the distal one according to the suggestions of Ball and Johnson [27] (Figure 1). EMG and goniometer signals were fed into an ADconverter with a 12 bit resolution. Raw signals were collected for one second with a sampling rate of 1000 Hz per channel and stored on a PC for subsequent analysis. Customized software was used for automatic detection of the following parameters: onset and termination of the tilting movement; onset of the foot inversion movement; inversion angle at the end of the tilting movement; maximum inversion angle; and activity onset for the four muscles (Figure 2). Reaction times (or latencies) of the four muscles were determined by the time difference between muscle activity onset and onset of the tilting movement. The average inversion velocity was calculated from the first onset of the inversion
34
D. ROSENBAUM ET AL.
Figure 2 Sample recording showing the four EMG and two angle signals with cursor positions for the respective parameters: o1, o2, o3 and o4 = activity onsets of the respective muscles; gA = onset of foot inversion movement; gB = inversion angle at the end of the tilting movement; gC = maximum inversion angle; ka/kB = onset and termination of the tilting movement.
2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
PERONEAL REACTION TIMES
35
Figure 3 Scattergram of all peroneus longus and brevis (PL/PB) reaction times for the stable and unstable (s/u) side of the mechanically and functionally unstable (m/f) subjects.
movement to the termination of the tilting movement using the value of the respective inversion angle. The program suggested a cursor position for each parameter that could be manually corrected in those cases in which the respective algorithm did not find the obviously correct position (e.g. in case of increased background muscle activity). The parameter values for each trial were stored and subsequently averaged for the six trials in each condition. Individual mean values were calculated from the six separate trials under each condition and used to determine mean values for the experimental groups. Due to the non-normal distribution of the data nonparametric procedures were applied. Differences between the two experimental groups were tested for significance by means of a Mann–Whitney U-test performed for each parameter. A Wilcoxon signed rank test was used to evaluate intraindividual differences. A significance level of P < 0.05 was chosen.
Results The two experimental groups were comparable with respect to age, height and body mass (Table 1). The 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
Table 1 Mean values (range) of the anthropometric and clinical data for the two experimental groups Functional instability (n=35) Age (years) 23.6 (19–37) Height (cm) 178.6 (165–189) Body mass (kg) 79.9 (55–110) Inversion trauma (per year) 7.9 (1–28) Radiological examinations Talar tilt Stable leg 2.4. (0–13) Unstable leg 6.4 (0–24) Difference 2.1 (−3–10) Anterior drawer (mm) Stable leg 5.7 (3–10) Unstable leg 6.8 (4–12) Difference 0.9 (-2–8)
Mechanical instability (n=30) 24.8 179.0 77.7 13.3
(19–38) (162–200) (56–104) (1–40)
3.9 (0–12) 7.6 (0–20) 4.8 (−6–18) 6.0 (3–10) 7.3 (4–12) 1.5 (-1–6)
group with MI sustained more inversion injuries/ month than the FI group (7.9 versus 13.3 injuries/ year). The clinical results revealed a more pronounced joint instability in the MI group that was reflected in greater values for the talar tilt and anterior drawer measurements of the stable and unstable limb as well as for the intraindividual differences (Table 1). The fact that the differences were less pronounced
36
D. ROSENBAUM ET AL.
Table 2 Mean values (range) of kinematic and electromyographic parameters for the experimental groups
Peroneus brevis latency (ms) Stable leg Unstable leg Difference Peroneus longus latency (ms) Stable leg Unstable leg Difference Tilt duration (ms) Stable leg Unstable leg Inversion angle (°) Stable leg Unstable leg Inversion velocity (°/s) Stable leg Unstable leg
Functional instability (n=35)
Mechanical instability (n=30)
Significance level (between groups)
57.5 (36–72) 64.7 (51–77) 7.7 (0–34)
63.0 (47–75) 57.2 (42–73) −5.7 (-17–0)
0.02 0.002 <0.0001
54.4. (38–75) 60.2 (38–77) 5.8 (−9–22)
57.9 (40–71) 53.6 (38–71) −4.3 (−18–8)
ns 0.01 <0.0001
57.7 (44–69) 59.1 (48–76)
58.1 (50–76) 58.3 (50–70)
ns ns
22.8 (11–33) 22.9 (13–31)
25.4 (19–34) 24.1 (16–35)
ns ns
326 (205–541) 301 (135–465)
ns
320 (137–506) 304 (151–444)
ns = not significant; P>0.05.
in the FI group suggests that in these subjects the instability problem cannot be explained only by a mechanical insufficiency of the lateral ankle ligaments. The experimental set-up led to a 30° trapdoor movement which lasted less than 60 ms (corresponding to an average angular velocity of 500 °/s, Table 2). This movement was fast enough to terminate before or at the time of the first muscle onsets (Figure 2). The values for the stable and unstable leg were similar in both groups. This indicates that the subjects loaded the tilting leg equally on both sides and did not try to protect the unstable leg by loading it less. For the inversion angle as well as for the inversion velocity no significant differences between the two experimental groups were found. Only a slight trend towards higher inversion velocities could be observed in the FI subjects. The peroneus longus was activated first, followed closely by the peroneus brevis with a delay of ≈3–4 ms. In both groups the peroneal muscles of the stable leg reacted with similar latencies. In the unstable limb the FI group revealed significantly longer latencies than the MI group. The difference was 7.5 ms for the peroneus brevis (P = 0.002) and 6.6 ms for the peroneus longus (P = 0.01). In the MI
group the peroneal muscles reacted earlier in the unstable than in the stable leg whereas in the FI group they reacted later. These intaindividual differences were highly significant (P < 0.0001). The tibialis anterior and the soleus were generally activated later than the peroneal muscles. The latencies were considerably more variable within a subject when compared to the peroneal muscles. Therefore, the activation of these muscles did not appear to be directly related to the inversion event and were not further analysed.
Discussion The study demonstrated the significant difference of peroneal reaction times between patients suffering from mechanical or functional instability after they had been separated based on conventional clinical diagnostics. This supports the view of functional instability as a neuromuscular problem so that the term ‘neuromuscular instability’ appears to describe it more precisely. Freeman [16] suggested that functional instability may be the result of impaired co-ordination following articular de-afferentiation, and that it interferes with the reflexes that depend on articular mechanoreceptors. 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
PERONEAL REACTION TIMES
The values obtained for the peroneal reaction times compared well with previous investigations even though the differences between functional and mechanical instability (8 ms for the peroneus brevis, 6 ms for the peroneus longus) were not as distinct in the present study as those reported before. Konradsen and Ravn [11] found significantly longer reaction times in athletes with functional instability compared to controls. For the peroneus brevis the difference was 15 ms, for the peroneus longus 17 ms. Johnson and Johnson [8] found latency differences of 8 ms in a surgically treated group as compared to 7 ms in normal subjects. Significant intraindividual differences of 16 ms for the peroneus longus and 13 ms for the peroneus brevis were observed in 20 athletes with unilateral ankle instability [9] while Nawoczenski et al. [14] found intraindividual differences of 9 ms in normal subjects and 14 ms in injured subjects. These intraindividual differences could not be confirmed by the present results in either experimental group. Even though these results differ slightly they all support the hypothesis that damage to the ankle may be reflected in impaired peroneal muscle function. The clinical and functional relevance of possible latency differences has been questioned. Based on experimental evidence and theoretical considerations, Thonnard et al. [17] claimed that the time necessary to reach the angle of capsular-ligamentous rupture is inferior to the observed reflex latencies and concluded that ankle sprains can only be avoided by anticipated active muscle stiffness. For the knee joint, on the other hand, it has been speculated that ‘the receptors in the knee joint ligaments probably contribute, via the gamma-muscle-spindle system, to preparatory adjustment (presetting) of the stiffness of the muscles around the knee joint, and thereby to the joint stiffness and the functional joint stability.’ [18]. More recent results indicate that muscle timing and recruitment order in response to anterior tibial translation are affected by anterior cruciate ligament injury [19–22]. Similar mechanisms might be applicable with respect to lateral ankle instability. Even if the reactions of peroneal muscles are not markedly different they may still indicate a different presetting of the muscles with the potential to prevent or control excessive ankle joint inversion. With respect to the kinematics of dynamic inversion movements we were surprised to see that 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
37
they did not reveal characteristics of the joint stability similar to those seen under static loading conditions of the radiographic and mechanical stress tests. A possible explanation would be that under the given conditions the mechanically unstable subjects used a faster muscle reaction to prevent excessive joint movements. It demonstrates, however, that in the mechanically unstable subjects the feedback mechanism operated fast enough to produce the necessary muscular response for joint stabilization at least in this experimental situation with an expected inversion event. Some limitations of the present study design should be mentioned. The experimental set-up did not expose the subjects to the same situation that they would encounter in real life. They were informed about the upcoming event and were mentally prepared to react appropriately. Therefore, the given situation was only a model of a real inversion event. In this context the differences in peroneal reaction times gain even more importance since they persisted in spite of the subject’s awareness of the potentially dangerous situation. Thus, subjects with functional instability could not benefit from a presetting of the muscles in order to reduce muscle latencies. Another point of discussion are the stress radiograph examinations of the ankle joint that are still the standard tool for determination of the extent of ligamentous lesion [23]. In these tests, the anterior drawer sign and talar tilt are determined from the radiograph image using absolute values or intraindividual differences between injured and healthy limb as criteria. However, there still is some disagreement about the acceptable limits and no general criteria exist to distinguish between the injury of one or more ligaments [24]. Furthermore, the accuracy and the specificity of the standard radiographic stress tests have been questioned since the results are dependent on the foot position, the applied force and the patient’s resistance [25]. Stress radiographs have been shown to underestimate the true talar tilt or anterior drawer due to reflexive muscle activity and should ideally be applied only under local anaesthesia [25, 26]. Although radiographic stress diagnostics are not considered to be painful for patients with chronic instability, it remains unclear whether a reflexive contraction of the muscles across the ankle joint may have an influence on the results. These standard clinical
38
D. ROSENBAUM ET AL.
procedures may be sufficient to characterize patients with extreme joint instability. In clinical practice, however, many patients cannot clearly be attributed as either functionally or mechanically unstable. This is even more difficult since functional instability, according to Tropp, implies motion beyond voluntary control without necessarily exceeding physiological range of motion. It is difficult to diagnose patients preoperatively and decide between different treatment options. Surgical treatment can be considered as over-therapy in the case of functional neuromuscular instability. Therefore, the main incentive for our study was to identify those patients that would benefit from physical therapy vs. those that would need surgery. Thereby, we would hope to avoid unnecessary operations in order to reduce patient discomfort and costs for the health care system. The presented experiments provided information about the functional capabilities of the peroneal muscles that was valuable for the differentiation of mechanical and functional neuromuscular instability even though the grouping was based on the above mentioned critical procedures. It still has to be discussed whether the applied procedures are appropriate to be used as standard clinical procedures. At this point, the experimental set-up and procedure are time consuming and need expert knowledge. More experience is necessary before standard values can be offered.
Acknowledgements This project was partly supported by the FraunhoferGesellschaft (Grant no. InSan I 1089-V-9092) and the Deutsche Forschungsgemeinschaft (DFG Grant no. CL77/3–1). We would like to thank Herbert Schmitt for the construction of the tilting platform and Michael Gold for the data analysis program.
References 1 Karlsson J, Lansinger O. Lateral instability of the ankle joint. Clin Orthop 1992; 276: 253–261. 2 Nawoczenski DA, Owen MG, Ecker ML et al. Objective evaluation of peroneal response to sudden inversion stress. J Orthop Sports Phys Ther 1985; 7: 107–109. 3 Thonnard JL, Plaghki L, Willems P et al. La pathoge´nie de l’entorse de la cheville: test d’une hypothe´se. Med Physica 1986; 9: 141–146.
4 Karlsson J, Lansinger O. Chronic lateral instability of the ankle in athletes. Sports Med 1993; 16: 355–365. 5 Freeman MAR. Instability of the foot after injuries to the lateral ligaments of the ankle. J Bone Joint Surg 1965; 47B: 669–677. 6 Gollhofer A, Scheuffelen C, Lohrer H. Neuromuskula¨re Stabilisation im oberen Sprunggelenk nach Immobilisation. Sportverl Sportschad 1993; 7: 23–28. 7 Isakov E, Mizrahi J, Solzi P et al. Response of peroneal muscles to sudden inversion of the ankle during standing. Int J Sports Biomech 1986; 2: 100–109. 8 Johnson MB, Johnson CL. Electromyographic response of peroneal muscles in surgical and nonsurgical injured ankles during sudden inversion. J Orthop Sports Phys Ther 1993; 18: 497–501. 9 Karlsson J, Peterson L, Andreasson G et al. Functional instability in mechanical unstable ankle joint. Acta Orthop Scand 1988; 59: 70–71. 10 Kimura IF, Nawoczenski DA, Epler M et al. Effect of the air stirrup in controlling ankle inversion stress. J Orthop Sports Phys Ther 1987; 9: 190–193. 11 Konradsen L, Bohsen Ravn J. Prolonged peroneal reaction time in ankle instability. Int J Sports Med 1991; 13: 290–292. 12 Larsen E, Lund PM. Peroneal muscle function in chronically unstable ankles: a prospective and postoperative electromyographic study. Clin Orthop 1991; 272: 219–226. 13 Marder RA. Current methods for the evaluation of ankle ligament injuries. J Bone Joint Surg 1994; 76A: 1103–1111. 14 Mizrahi J, Ramot Y, Susak Z. The dynamics of the subtalar joint in sudden inversion of the foot. J Biomech Eng 1990; 112: 9–14. 15 Lo¨fvenberg R, Ka¨rrholm J, Sundelin G et al. Prolonged reaction time in patients with chronic lateral instability of the ankle. Am J Sports Med 1995; 23: 414–417. 16 Freeman MAR, Dean MRE, Hanham IWF. The etiology and prevention of functional instability of the foot. J Bone Joint Surg 1965; 47B: 678–685. 17 Rijke AM, Jones B, Vierhout PAM. Stress examination of traumatized lateral ligaments of the ankle. Clin Orthop 1986; 210: 143–151. 18 Johansson H, Sjo¨la¨nder P, Sojka P. Receptors in the knee joint ligaments and their role in the biomechanics of the joint. Crit Rev Biomed Eng 1991; 18: 341–368. 19 Tropp H, Odenrick P. Postural control in single-limb stance. J Orthop Res 1988; 6: 833–839. 20 Tropp H, Odenrick P, Gillquist J. Stabilometry recordings in functional and mechanical instability of the ankle joint. Int J Sports Med 1985; 6: 180–182. 21 Tropp H, Ekstrand J, Gillquist J. Factors affecting stabilometry recordings of single limb stance. Am J Sports Med 1984; 12: 185–188. 22 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. 23 Lucht U, Vang PS, Termansen NB. Lateral ligament reconstruction of the ankle with a modified Watson–Jones operation. Acta Orthop Scand 1981; 52: 363–399. 24 Chandnani VP, Harper MT, Ficke JR et al. Chronic ankle instability: evaluation with MR arthrography, MR imaging, and stress radiography. Radiology 1994; 192: 189–194. 25 Peters JW, Trevino SG, Renstrøm PA. Chronic lateral ankle instability. Foot Ankle 1991; 12: 182–191. 26 Becker H-P, Komischke A, Danz B et al. Stress diagnostics of the sprained ankle: evaluation of the anterior drawer test with and without anesthesia. Foot Ankle 1993; 14: 459–464. 27 Ball P, Johnson GR. Reliability of hindfoot geometry when using a flexible electrogoniometer. Clin Biomech 1993; 8: 13–19.
2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
6: 31–38
Peroneal reaction times for diagnosis of functional ankle instability D. ROSENBAUM,∗ H.-P. BECKER,† H. GERNGRO߆ AND L. CLAES‡ ∗Motion Analysis Laboratory, Orthopaedic Department, University of Mu¨nster, Mu¨nster, †Surgical Department, Military Hospital, Ulm and ‡Department of Orthopaedic Research and Biomechanics, University of Ulm, Ulm, Germany
Summary In order to differentiate between functional and mechanical chronic ankle instability, peroneal muscle function was examined in sudden inversion movements produced by a 30° tilting platform. Surface electromyographic (EMG) signals were recorded from the peroneus brevis and longus, tibialis anterior, and soleus muscles. An electrogoniometer was used to record rearfoot motion. Sixty-five subjects were assigned either to a functional (n = 35) or a mechanical instability group (n = 30) based on radiographic and mechanical stability measurements. Mechanical instability was characterized by a greater talar tilt and anterior drawer in the unstable leg. The groups did not differ significantly in the stability characteristics of the unaffected leg. Functional instability was reflected in significantly longer reaction times of the peroneus brevis (65 versus 57 ms, P = 0.002) and longus muscle (60 versus 54 ms, P = 0.01) in comparison with the mechanical instability group. Therefore, in those unstable patients without a clear mechanical insufficiency, the problem appears to be caused by a neuromuscular instability. Peroneal reaction times may help to distinguish between different causes of chronic ankle instability. Keywords: chronic ankle instability; peroneal reaction time; surface electromyography; tilting platform
Introduction Despite adequate treatment of acute lateral ankle ligament injuries, chronic instability–mechanical or functional–occurs in ≈10–20% of the patients [1, 2]. Mechanical instability is caused by a ligament Correspondence: Priv. Doz. Dr Dieter Rosenbaum, Funktionsbereich Bewegungsanalytik, Klinik und Poliklinik fu¨r Allgemeine Orthopa¨die, Westfa¨lische Wilhelms-Universita¨t Mu¨nster, Albert-Schweitzer-Str. 33, D-48129 Mu¨nster, Germany (email: [email protected]). 2000 Blackwell Science Ltd
deficiency and refers objectively to clinical and radiographic stress tests. Functional instability comprises the patient’s subjective feeling of repeated giving way based on proprioceptive disorders or muscle weakness [2, 3]. However, it is often difficult to distinguish between these two entities because a mechanical instability may arise from functional instability. Recurrent inversion injury of the ankle may be related either to mechanical or functional factors alone or to both. Since they are inconsistently associated, it is clinically difficult to decide whether
31
32
D. ROSENBAUM ET AL.
surgery is needed or not, especially in cases with borderline radiographic findings. Presently, no objective measure exists for the diagnosis of functional instability and it is generally dependent on the physician’s judgement alone. Before deciding upon a surgical reconstruction in a patient with chronic instability, a rehabilitation program is implemented. Approximately 50% of the patients will regain satisfactory functional stability after 12 weeks [4]. The remaining 50% will lose valuable 3 months with the risk of new injuries. Therefore, it is mandatory to seek useful tests for the diagnosis of functional instability and proprioceptive disorders. Freeman et al.[5] were the first to report on functional ankle instability. They confirmed the diagnosis using the clinical Rhomberg-test and demonstrated that strengthening and co-ordination exercises were able to relieve the symptoms. In athletes, Tropp found muscle weakness and an impaired ability to maintain postural equilibrium to be associated with functional instability [3]. Recently, trapdoor measurements have been used to investigate the reaction time of the peroneal muscles to an inversion movement [3,6–14]. Some of these studies demonstrated a generally slower reaction time in unstable ankles [11, 14, 15]. There is no study, however, which uses the information on muscle function for the diagnosis of functional instability and subsequent clinical decision making. The aim of the present study was to evaluate whether reaction time of the peroneal muscles provoked by sudden inversion movement is a useful parameter to distinguish between mechanical and functional instability.
Materials and methods Sixty-five patients with self-reported lateral ankle instability volunteered to participate in this prospective study. Thirty-five patients had been excluded due to bilateral symptoms, hyper-mobility and severe pain problems. A detailed questionnaire was used to evaluate the patients’ history of instability and the frequency of inversion traumata. Clinical and radiographic examinations of the ankle joint complex were analysed by the same physician (H.P.B.). Stress radiographs measuring anterior drawer and talar tilt were performed using a Scheubaapparatus (Telos Co., Hungen, Germany).
Based on the results of the radiographic and mechanical stress analyses, the subjects were separated into one of the following groups: (1) MI = mechanical instability (n = 30); and (2) FI = functional instability (n = 35). Three of the following criteria had to be fulfilled for a subject to qualify for the MI group:
1. More than eight inversion traumata per month. 2. Joint instability in the manual drawer and inversion test clearly increased in comparison to the contralateral side. 3. Radiographic talar tilt >7° and the difference between unstable and contralateral ankle >0°; 4. Radiographic anterior drawer >7 mm and the difference between unstable and contralateral ankle >0 mm. In order to evaluate the diagnostic value of the applied procedure the ligament status of those mechanically unstable patients (n = 26) who agreed to surgical treatment was determined intraoperatively. The same surgeon inspected the lateral ligaments (ATFL and CFL) and described it either as stable (intact fibrous structures or strong scar tissue) or unstable (elongated or defect). The findings were compared with the results of the reaction time measurements. A tilting platform with a mechanical release mechanism was built (Figure 1). The platform was 20 cm high and had a 44×40 cm wooden surface. The tilting part was 19 cm wide and was mounted on roller bearings. An aluminium lath was aligned with the lateral edge of the foot in order to prevent slipping. The trapdoor was released by pushing a foot switch not visible for the subject. Upon release the platform fell down through an arch of 30° which was predetermined by a mechanical stop (maximum angle of 45°). A potentiometer aligned with the axis of rotation was used to record the onset and duration of the tilting movement. After preparation the subjects were asked to stand on the titling platform. Six dynamic inversion measurements were executed in a random order in both extremities. In order to simulate an inversion trauma more realistically the patients were asked to shift the body weight onto the tilting foot with the contralateral foot only slightly planted to keep balance, i.e. almost full body weight loading. 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
PERONEAL REACTION TIMES
33
Figure 1 Schematic of the experimental set-up with a rear view of the subject’s lower legs and the electrogoniometer fixed to the heel and calf in line with the longitudinal leg axis.
Electromyographic (EMG) signals were recorded from four lower leg muscles – i.e. the peroneus longus, peroneus brevis, tibialis anterior and soleus – by means of custom made surface electrodes with on-site preamplification (gain factor 10). The signals were fed into a battery operated amplifier with a gain factor of 100, an input impedance over 1010 Ohm, a frequency range from 15 to 1056 Hz, and a common mode rejection ratio of 84 db. The electrodes had two silver detection surfaces with a diameter of 12 mm and an interelectrode distance of 20 mm centre-to-centre. In order to reduce skin impedance the recording sites were prepared by shaving, cleaning with 70% alcohol, and applying electrode gel. The electrodes were fixed with double-adhesive washers on the skin over the respective muscles. The muscles were palpated during contraction. The peroneus longus electrode was located about 3 cm below the proximal head of the fibula, the peroneus brevis electrode about 5 cm above the lateral malleolus. The tibialis anterior electrode was fixed 1 cm lateral to the tibial edge, the soleus electrode right below the head of the gastrocnemius near the peroneus brevis electrode. After preparation the electrode fixation and signal quality were controlled for movement and cable artifacts by tapping against the electrode. Correct positioning was controlled by asking the subject to perform isolated foot movements in order to elicit activity from the 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
respective muscles – peroneus brevis and longus: eversion movements with the foot planted; tibialis anterior: lifting the forefoot; soleus: raising on the toes. This procedure also allowed for a qualitative control of crosstalk artifacts since during isolated muscle activity in one EMG channel the other channels should remain silent. Foot eversion/inversion measurements were recorded with a strain-gauge goniometer (Type M 110, Penny & Giles, Blackwood, UK). The distal end block was taped to the calcaneus and the proximal end block was taped to the calf in line with the distal one according to the suggestions of Ball and Johnson [27] (Figure 1). EMG and goniometer signals were fed into an ADconverter with a 12 bit resolution. Raw signals were collected for one second with a sampling rate of 1000 Hz per channel and stored on a PC for subsequent analysis. Customized software was used for automatic detection of the following parameters: onset and termination of the tilting movement; onset of the foot inversion movement; inversion angle at the end of the tilting movement; maximum inversion angle; and activity onset for the four muscles (Figure 2). Reaction times (or latencies) of the four muscles were determined by the time difference between muscle activity onset and onset of the tilting movement. The average inversion velocity was calculated from the first onset of the inversion
34
D. ROSENBAUM ET AL.
Figure 2 Sample recording showing the four EMG and two angle signals with cursor positions for the respective parameters: o1, o2, o3 and o4 = activity onsets of the respective muscles; gA = onset of foot inversion movement; gB = inversion angle at the end of the tilting movement; gC = maximum inversion angle; ka/kB = onset and termination of the tilting movement.
2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
PERONEAL REACTION TIMES
35
Figure 3 Scattergram of all peroneus longus and brevis (PL/PB) reaction times for the stable and unstable (s/u) side of the mechanically and functionally unstable (m/f) subjects.
movement to the termination of the tilting movement using the value of the respective inversion angle. The program suggested a cursor position for each parameter that could be manually corrected in those cases in which the respective algorithm did not find the obviously correct position (e.g. in case of increased background muscle activity). The parameter values for each trial were stored and subsequently averaged for the six trials in each condition. Individual mean values were calculated from the six separate trials under each condition and used to determine mean values for the experimental groups. Due to the non-normal distribution of the data nonparametric procedures were applied. Differences between the two experimental groups were tested for significance by means of a Mann–Whitney U-test performed for each parameter. A Wilcoxon signed rank test was used to evaluate intraindividual differences. A significance level of P < 0.05 was chosen.
Results The two experimental groups were comparable with respect to age, height and body mass (Table 1). The 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
Table 1 Mean values (range) of the anthropometric and clinical data for the two experimental groups Functional instability (n=35) Age (years) 23.6 (19–37) Height (cm) 178.6 (165–189) Body mass (kg) 79.9 (55–110) Inversion trauma (per year) 7.9 (1–28) Radiological examinations Talar tilt Stable leg 2.4. (0–13) Unstable leg 6.4 (0–24) Difference 2.1 (−3–10) Anterior drawer (mm) Stable leg 5.7 (3–10) Unstable leg 6.8 (4–12) Difference 0.9 (-2–8)
Mechanical instability (n=30) 24.8 179.0 77.7 13.3
(19–38) (162–200) (56–104) (1–40)
3.9 (0–12) 7.6 (0–20) 4.8 (−6–18) 6.0 (3–10) 7.3 (4–12) 1.5 (-1–6)
group with MI sustained more inversion injuries/ month than the FI group (7.9 versus 13.3 injuries/ year). The clinical results revealed a more pronounced joint instability in the MI group that was reflected in greater values for the talar tilt and anterior drawer measurements of the stable and unstable limb as well as for the intraindividual differences (Table 1). The fact that the differences were less pronounced
36
D. ROSENBAUM ET AL.
Table 2 Mean values (range) of kinematic and electromyographic parameters for the experimental groups
Peroneus brevis latency (ms) Stable leg Unstable leg Difference Peroneus longus latency (ms) Stable leg Unstable leg Difference Tilt duration (ms) Stable leg Unstable leg Inversion angle (°) Stable leg Unstable leg Inversion velocity (°/s) Stable leg Unstable leg
Functional instability (n=35)
Mechanical instability (n=30)
Significance level (between groups)
57.5 (36–72) 64.7 (51–77) 7.7 (0–34)
63.0 (47–75) 57.2 (42–73) −5.7 (-17–0)
0.02 0.002 <0.0001
54.4. (38–75) 60.2 (38–77) 5.8 (−9–22)
57.9 (40–71) 53.6 (38–71) −4.3 (−18–8)
ns 0.01 <0.0001
57.7 (44–69) 59.1 (48–76)
58.1 (50–76) 58.3 (50–70)
ns ns
22.8 (11–33) 22.9 (13–31)
25.4 (19–34) 24.1 (16–35)
ns ns
326 (205–541) 301 (135–465)
ns
320 (137–506) 304 (151–444)
ns = not significant; P>0.05.
in the FI group suggests that in these subjects the instability problem cannot be explained only by a mechanical insufficiency of the lateral ankle ligaments. The experimental set-up led to a 30° trapdoor movement which lasted less than 60 ms (corresponding to an average angular velocity of 500 °/s, Table 2). This movement was fast enough to terminate before or at the time of the first muscle onsets (Figure 2). The values for the stable and unstable leg were similar in both groups. This indicates that the subjects loaded the tilting leg equally on both sides and did not try to protect the unstable leg by loading it less. For the inversion angle as well as for the inversion velocity no significant differences between the two experimental groups were found. Only a slight trend towards higher inversion velocities could be observed in the FI subjects. The peroneus longus was activated first, followed closely by the peroneus brevis with a delay of ≈3–4 ms. In both groups the peroneal muscles of the stable leg reacted with similar latencies. In the unstable limb the FI group revealed significantly longer latencies than the MI group. The difference was 7.5 ms for the peroneus brevis (P = 0.002) and 6.6 ms for the peroneus longus (P = 0.01). In the MI
group the peroneal muscles reacted earlier in the unstable than in the stable leg whereas in the FI group they reacted later. These intaindividual differences were highly significant (P < 0.0001). The tibialis anterior and the soleus were generally activated later than the peroneal muscles. The latencies were considerably more variable within a subject when compared to the peroneal muscles. Therefore, the activation of these muscles did not appear to be directly related to the inversion event and were not further analysed.
Discussion The study demonstrated the significant difference of peroneal reaction times between patients suffering from mechanical or functional instability after they had been separated based on conventional clinical diagnostics. This supports the view of functional instability as a neuromuscular problem so that the term ‘neuromuscular instability’ appears to describe it more precisely. Freeman [16] suggested that functional instability may be the result of impaired co-ordination following articular de-afferentiation, and that it interferes with the reflexes that depend on articular mechanoreceptors. 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
PERONEAL REACTION TIMES
The values obtained for the peroneal reaction times compared well with previous investigations even though the differences between functional and mechanical instability (8 ms for the peroneus brevis, 6 ms for the peroneus longus) were not as distinct in the present study as those reported before. Konradsen and Ravn [11] found significantly longer reaction times in athletes with functional instability compared to controls. For the peroneus brevis the difference was 15 ms, for the peroneus longus 17 ms. Johnson and Johnson [8] found latency differences of 8 ms in a surgically treated group as compared to 7 ms in normal subjects. Significant intraindividual differences of 16 ms for the peroneus longus and 13 ms for the peroneus brevis were observed in 20 athletes with unilateral ankle instability [9] while Nawoczenski et al. [14] found intraindividual differences of 9 ms in normal subjects and 14 ms in injured subjects. These intraindividual differences could not be confirmed by the present results in either experimental group. Even though these results differ slightly they all support the hypothesis that damage to the ankle may be reflected in impaired peroneal muscle function. The clinical and functional relevance of possible latency differences has been questioned. Based on experimental evidence and theoretical considerations, Thonnard et al. [17] claimed that the time necessary to reach the angle of capsular-ligamentous rupture is inferior to the observed reflex latencies and concluded that ankle sprains can only be avoided by anticipated active muscle stiffness. For the knee joint, on the other hand, it has been speculated that ‘the receptors in the knee joint ligaments probably contribute, via the gamma-muscle-spindle system, to preparatory adjustment (presetting) of the stiffness of the muscles around the knee joint, and thereby to the joint stiffness and the functional joint stability.’ [18]. More recent results indicate that muscle timing and recruitment order in response to anterior tibial translation are affected by anterior cruciate ligament injury [19–22]. Similar mechanisms might be applicable with respect to lateral ankle instability. Even if the reactions of peroneal muscles are not markedly different they may still indicate a different presetting of the muscles with the potential to prevent or control excessive ankle joint inversion. With respect to the kinematics of dynamic inversion movements we were surprised to see that 2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38
37
they did not reveal characteristics of the joint stability similar to those seen under static loading conditions of the radiographic and mechanical stress tests. A possible explanation would be that under the given conditions the mechanically unstable subjects used a faster muscle reaction to prevent excessive joint movements. It demonstrates, however, that in the mechanically unstable subjects the feedback mechanism operated fast enough to produce the necessary muscular response for joint stabilization at least in this experimental situation with an expected inversion event. Some limitations of the present study design should be mentioned. The experimental set-up did not expose the subjects to the same situation that they would encounter in real life. They were informed about the upcoming event and were mentally prepared to react appropriately. Therefore, the given situation was only a model of a real inversion event. In this context the differences in peroneal reaction times gain even more importance since they persisted in spite of the subject’s awareness of the potentially dangerous situation. Thus, subjects with functional instability could not benefit from a presetting of the muscles in order to reduce muscle latencies. Another point of discussion are the stress radiograph examinations of the ankle joint that are still the standard tool for determination of the extent of ligamentous lesion [23]. In these tests, the anterior drawer sign and talar tilt are determined from the radiograph image using absolute values or intraindividual differences between injured and healthy limb as criteria. However, there still is some disagreement about the acceptable limits and no general criteria exist to distinguish between the injury of one or more ligaments [24]. Furthermore, the accuracy and the specificity of the standard radiographic stress tests have been questioned since the results are dependent on the foot position, the applied force and the patient’s resistance [25]. Stress radiographs have been shown to underestimate the true talar tilt or anterior drawer due to reflexive muscle activity and should ideally be applied only under local anaesthesia [25, 26]. Although radiographic stress diagnostics are not considered to be painful for patients with chronic instability, it remains unclear whether a reflexive contraction of the muscles across the ankle joint may have an influence on the results. These standard clinical
38
D. ROSENBAUM ET AL.
procedures may be sufficient to characterize patients with extreme joint instability. In clinical practice, however, many patients cannot clearly be attributed as either functionally or mechanically unstable. This is even more difficult since functional instability, according to Tropp, implies motion beyond voluntary control without necessarily exceeding physiological range of motion. It is difficult to diagnose patients preoperatively and decide between different treatment options. Surgical treatment can be considered as over-therapy in the case of functional neuromuscular instability. Therefore, the main incentive for our study was to identify those patients that would benefit from physical therapy vs. those that would need surgery. Thereby, we would hope to avoid unnecessary operations in order to reduce patient discomfort and costs for the health care system. The presented experiments provided information about the functional capabilities of the peroneal muscles that was valuable for the differentiation of mechanical and functional neuromuscular instability even though the grouping was based on the above mentioned critical procedures. It still has to be discussed whether the applied procedures are appropriate to be used as standard clinical procedures. At this point, the experimental set-up and procedure are time consuming and need expert knowledge. More experience is necessary before standard values can be offered.
Acknowledgements This project was partly supported by the FraunhoferGesellschaft (Grant no. InSan I 1089-V-9092) and the Deutsche Forschungsgemeinschaft (DFG Grant no. CL77/3–1). We would like to thank Herbert Schmitt for the construction of the tilting platform and Michael Gold for the data analysis program.
References 1 Karlsson J, Lansinger O. Lateral instability of the ankle joint. Clin Orthop 1992; 276: 253–261. 2 Nawoczenski DA, Owen MG, Ecker ML et al. Objective evaluation of peroneal response to sudden inversion stress. J Orthop Sports Phys Ther 1985; 7: 107–109. 3 Thonnard JL, Plaghki L, Willems P et al. La pathoge´nie de l’entorse de la cheville: test d’une hypothe´se. Med Physica 1986; 9: 141–146.
4 Karlsson J, Lansinger O. Chronic lateral instability of the ankle in athletes. Sports Med 1993; 16: 355–365. 5 Freeman MAR. Instability of the foot after injuries to the lateral ligaments of the ankle. J Bone Joint Surg 1965; 47B: 669–677. 6 Gollhofer A, Scheuffelen C, Lohrer H. Neuromuskula¨re Stabilisation im oberen Sprunggelenk nach Immobilisation. Sportverl Sportschad 1993; 7: 23–28. 7 Isakov E, Mizrahi J, Solzi P et al. Response of peroneal muscles to sudden inversion of the ankle during standing. Int J Sports Biomech 1986; 2: 100–109. 8 Johnson MB, Johnson CL. Electromyographic response of peroneal muscles in surgical and nonsurgical injured ankles during sudden inversion. J Orthop Sports Phys Ther 1993; 18: 497–501. 9 Karlsson J, Peterson L, Andreasson G et al. Functional instability in mechanical unstable ankle joint. Acta Orthop Scand 1988; 59: 70–71. 10 Kimura IF, Nawoczenski DA, Epler M et al. Effect of the air stirrup in controlling ankle inversion stress. J Orthop Sports Phys Ther 1987; 9: 190–193. 11 Konradsen L, Bohsen Ravn J. Prolonged peroneal reaction time in ankle instability. Int J Sports Med 1991; 13: 290–292. 12 Larsen E, Lund PM. Peroneal muscle function in chronically unstable ankles: a prospective and postoperative electromyographic study. Clin Orthop 1991; 272: 219–226. 13 Marder RA. Current methods for the evaluation of ankle ligament injuries. J Bone Joint Surg 1994; 76A: 1103–1111. 14 Mizrahi J, Ramot Y, Susak Z. The dynamics of the subtalar joint in sudden inversion of the foot. J Biomech Eng 1990; 112: 9–14. 15 Lo¨fvenberg R, Ka¨rrholm J, Sundelin G et al. Prolonged reaction time in patients with chronic lateral instability of the ankle. Am J Sports Med 1995; 23: 414–417. 16 Freeman MAR, Dean MRE, Hanham IWF. The etiology and prevention of functional instability of the foot. J Bone Joint Surg 1965; 47B: 678–685. 17 Rijke AM, Jones B, Vierhout PAM. Stress examination of traumatized lateral ligaments of the ankle. Clin Orthop 1986; 210: 143–151. 18 Johansson H, Sjo¨la¨nder P, Sojka P. Receptors in the knee joint ligaments and their role in the biomechanics of the joint. Crit Rev Biomed Eng 1991; 18: 341–368. 19 Tropp H, Odenrick P. Postural control in single-limb stance. J Orthop Res 1988; 6: 833–839. 20 Tropp H, Odenrick P, Gillquist J. Stabilometry recordings in functional and mechanical instability of the ankle joint. Int J Sports Med 1985; 6: 180–182. 21 Tropp H, Ekstrand J, Gillquist J. Factors affecting stabilometry recordings of single limb stance. Am J Sports Med 1984; 12: 185–188. 22 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. 23 Lucht U, Vang PS, Termansen NB. Lateral ligament reconstruction of the ankle with a modified Watson–Jones operation. Acta Orthop Scand 1981; 52: 363–399. 24 Chandnani VP, Harper MT, Ficke JR et al. Chronic ankle instability: evaluation with MR arthrography, MR imaging, and stress radiography. Radiology 1994; 192: 189–194. 25 Peters JW, Trevino SG, Renstrøm PA. Chronic lateral ankle instability. Foot Ankle 1991; 12: 182–191. 26 Becker H-P, Komischke A, Danz B et al. Stress diagnostics of the sprained ankle: evaluation of the anterior drawer test with and without anesthesia. Foot Ankle 1993; 14: 459–464. 27 Ball P, Johnson GR. Reliability of hindfoot geometry when using a flexible electrogoniometer. Clin Biomech 1993; 8: 13–19.
2000 Blackwell Science Ltd, Foot and Ankle Surgery, 6, 31–38