Share this article on:

Modulation of Prelanding Lower-Limb Muscle Responses in Athletes with Multiple Ankle Sprains


Medicine & Science in Sports & Exercise: October 2007 - Volume 39 - Issue 10 - p 1774-1783
doi: 10.1249/mss.0b013e3181343629

Purpose: The objective of this study was to investigate modulation in prelanding muscle responses and its associated impact force on landing from unexpected and self-initiated drops in male basketball players with a history of bilateral multiple ankle sprains (BMAS).

Methods: Prelanding EMG responses were recorded in four lower-limb muscles, together with the impact force on landing, while 20 healthy and 19 basketball players with BMAS performed unexpected, self-initiated drops from a height of 30 cm.

Results: Group differences were detected after self-initiated but not unexpected drops. Two main changes in prelanding EMG responses were observed in the injured basketball players during the self-initiated drops. First, tibialis anterior (TA) was activated significantly earlier in the injured group, whereas left tensor fascia latae appeared closer to the moment of landing (P < 0.025) than in the healthy players. Second, cocontraction indexes between left TA and peroneus longus, and left TA and medial gastrocnemius, were significantly greater in the injured than in the healthy players (P < 0.025). On landing, higher magnitude-of-impact forces were observed in the injured players on the right leg (by 23%, P = 0.012).

Conclusion: In basketball players with BMAS, modulation of prelanding muscle response latencies occurred in injured (ankle) and uninjured (hip) joints during self-initiated but not unexpected drops. Greater cocontraction index between the left ankle muscle pairs in preparation for landing from self-initiated drops, and a significantly higher magnitude of impact force in the right leg on landing, were observed in the injured players.

Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, CHINA

Address for correspondence: Christina W. Y. Hui-Chan, Ph.D., Chair Professor, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong; E-mail:

Submitted for publication July 2006.

Accepted for publication June 2007.

Ankle sprain is one of the most common sports injuries at all level of sports (2), accounting for 20% of all injuries. Even though 95% of injured athletes are able to return to sports participation within 6 wk, about 40% of them experience postural instability during activities such as jumping and hopping, for 6 months to 1 yr after injury (4,18). Proprioceptive deficits as a consequence of the initial trauma to the ligament has been proposed as one possible factor that might lead to reinjury of the lateral ligaments of the injured ankle (14). Indeed, our previous study has shown that the average errors in passive ankle reposition in plantarflexion were significantly increased from 0.9° in the healthy to 1.25° in the injured basketball players (16). These findings agree with those obtained during active repositioning in plantarflexion (19). In any case, joint-repositioning deficits in ankle sprains seem to be relatively small, even though the percent changes were 39% for passive repositioning in ankle plantarflexion (16) and 34% for active repositioning in ankle inversion (23).

Joint proprioception deficits aside, possible changes in latencies and magnitude of muscle responses, in particular, the peroneal longus and brevis, had been studied during stance in subjects with a history of ankle sprains using tilting platform protocols (7,13,24). Results from these studies, however, have produced controversial findings. Only 3 of 10 studies report a significant difference in the latency of the peroneal muscle responses to sudden ankle inversion as a result of ankle sprains (24). There were no significant differences in response latencies in the other seven studies (7,13). Note that ankle sprains normally occur when the ankle was excessively plantarflexed and inverted (26), such as landing on an opponent's foot (2). In addition, most injured subjects have feeling of instability during landing from jumps (18). Considering that basketball players have to repeatedly land from jumps during a game, it is interesting to note that no studies have so far been conducted to assess possible changes in lower-limb EMG in preparation for landing in the injured subjects, under unanticipated (unexpected) and anticipated (self-initiated) situations.

Prelanding muscle activities in the lower limb have been observed during unexpected (20,33) and self-initiated drops (15,20). The muscle activities after unexpected drops are considered reflexive in nature (33), but those coming after self-initiated drops are classified as anticipatory behavior in preparation for landing (31,38). Proprioceptive, vestibular, and visual inputs are thought to trigger and modulate the prelanding muscle activation in both unexpected (20) and self-initiated drops (31). In addition, joint position sense is also suggested to contribute to the cocontraction between the antagonistic muscle pair of an involved joint during a preplanned movement (39). Thus, deficits or alterations in ankle proprioceptive inputs could have modulated the timing and scaling of the prelanding muscle responses in patients with ankle sprains somewhat differently during unexpected and self-initiated drops. Note that the control of landing involves coordination of the whole body (32), in that the motor control system needs to simultaneously coordinate many muscles across the different joints that act synergistically to maintain balance. In this connection, hip abductor/adductor muscles are suggested to control the leg-pelvis stiffness for frontal-plane motion during quiet standing (40) and lateral translation during multidirectional perturbations (21). As mentioned before, ankle sprains normally occur when the ankle is excessively plantarflexed and inverted (26), which involves excessive movements in both the sagittal and frontal planes. To ensure proper balance on landing from self-initiated drops, injured players might need to adopt a strategy that could involve uninjured (hip) as well as injured (ankle) joints for landing. Indeed, an examination of possible modulation in the prelanding muscle activities at the ankle and hip muscles, as well as the cocontraction between antagonistic muscle pairs at the ankle, is essential for understanding possible abnormal control of landing in these injured players. Such findings would provide the insight needed for designing effective intervention strategies to prevent reoccurrence of similar injuries.

Prelanding muscle responses have been thought to be responsible for the control of the early deceleration phase associated with landing from self-initiated drops (31). McKinley and Pedotti (31) suggest that the goal of such prelanding activation is to maintain balance and to minimize the impact force on landing. Furthermore, a higher impact force applied to the ankle has been shown to have greater potential to cause an ankle sprain (1) and to contribute to osteoarthritic changes in the lower-limb joints (37). Hence, any possible changes in the magnitude of the impact force on landing in the injured players when compared with the healthy ones should be evaluated.

Based on the above considerations, the aims of this study were twofold: 1) to determine possible modulation in prelanding muscle activities in the ankle and hip, and 2) to assess possible changes in the impact force on landing from unexpected and self-initiated drops in basketball players with and without a history of multiple ankle sprains. It was hypothesized that there would be modulation in the prelanding muscle responses in the uninjured (hip) and injured (ankle) joints and an increase in the amount of the impact force during landing from self-initiated drops that triggered anticipatory responses, but not during landing from unexpected drops that triggered reflex responses, between basketball players who had and who had not sustained bilateral ankle sprains.

Back to Top | Article Outline


Subjects and study design.

Thirty-nine male university basketball players, aged between 19 and 26 yr, participated in this study. Twenty of them were healthy subjects, and the remaining 19 had bilateral multiple ankle sprains (BMAS). Both groups were similar in age (mean ± SD = 20.8 ± 1.6 and 20.4 ± 1.0 yr, respectively) and weight (mean ± SD = 68.5 ± 8.7 and 70.6 ± 8.0 kg, respectively), but the BMAS group was, on average, slightly but significantly taller than the healthy group (by 0.04 m, mean = 1.8 ± 0.0 and 1.8 ± 0.1 m, respectively; P = 0.012). They were recruited from the basketball teams of seven local universities, and all subjects had participated in competition for more than 5 yr. The healthy players had no history of ankle sprain during the past 2 yr, but the injured players had sustained at least two inversion ankle sprains in each ankle during the previous 2 yr. These ankle sprains had been severe enough for them to seek medical intervention. The time since the last ankle sprain was 6.0 ± 3.4 months. Thirteen of the 19 subjects (63%) in the BMAS group had feelings of postural instability, especially during landing from jumps. Players who had other musculoskeletal disorders, fracture or surgery to the lower limb or back were excluded. The study was approved by the ethical committee of the Hong Kong Polytechnic University, and all subject signed an informed consent before beginning the study.

Back to Top | Article Outline

Drop test.

There are two methods for testing prelanding muscle activities and impact force on landing: drop from a height and jump to the ground. In this study, the former test was adopted by dropping subjects from the ceiling in a safety harness. Care was taken to ensure that all tested muscles were relaxed before the subject was released from the electromagnet mounted on the ceiling, via visual monitoring of the EMG records on the computer screen. Drops were performed under two conditions (unexpected and self-initiated) in a random order with eyes open. Normalization with reference to maximum voluntary isometric contraction (MVIC) was commonly used. However, maximum capacity of a muscle during a dynamic task might exceed the muscle activities during MVIC (8), and normalization using other reference values such as the ensemble peak of a stride during walking (41); or the peak amplitude of a particular condition during gait analysis had been proposed (6). In this connection, we observed that lower-limb EMG activities were greater in response to unexpected drops than to MVIC and self-initiated drops (15). Consequently, prelanding TA, PER, and MG response amplitudes were normalized with both MVIC and the ensemble peak EMG activities recorded during the prelanding phase of unexpected drops for cross reference.

Back to Top | Article Outline


The subject was suspended in a safety harness that was attached to a metal plate mounted on the ceiling via a strong electromagnet, with his toes 30 cm above two separate forceplates (Kistler, Winterthur, Switzerland) (Fig. 1). The electromagnetic circuit was switched off by the experimenter to initiate unexpected drops, and by the subject to initiate self-initiated drops. A verbal warning signal was given before the experimenter initiated unexpected drops. To ensure that subjects could not predict the exact moment of drop, we had incorporated a time lag ranging from 5 to 20 s between the warning signal and the actual release. A small circuit connected to the magnet indicated the moment of release, and the moment of landing and the resulting impact forces were registered by a separate forceplate under each foot. Each subject had a total of 20 drops. Unexpected and self-initiated drops were conducted in blocks of 10, with the order of conditions randomized. They were requested to land on both legs and to maintain their balance on landing.



To record MVIC of the TA and PER muscles, the subject lied in supine on a Cybex Isokinetic dynamometer (Cybex Norm, Cybex International Inc., Ronkonkoma, NY), with his ankle joint kept in a neutral position (defined here as 90°of ankle dorsiflexion) and his knee joint in 90° of flexion. To test for the MG muscle, the subject lay prone, with his ankle joint maintained in a neutral position but with his knee joint in full extension. Before data collection, all subjects performed three submaximal contractions as practice trials, followed by three maximum isometric contractions of each muscle. When maximum force had reached, subjects were encouraged to maintain their maximum contraction for 6 s. About 10 s of EMG signals were collected for further analysis.

Back to Top | Article Outline

EMG recordings.

The skin surface overlying the muscles was shaved, cleaned with 70% alcohol, and rubbed with an abrasive gel to remove grease and dead skin from the surface. Surface electrodes (1.25 cm in diameter, with the center-to-center interelectrode distance of 2 cm, NeuroCom International, Inc) with built-in preamplifiers having a gain of 290 were placed parallel to the muscle fibers over the following four muscles: 1) tensor fascia latae (TFL), 2 cm below the anterior superior iliac spine, palpated with the leg extended and the subject in lying (5); 2) peroneus longus (PER), one third of the distance between the fibular head and the lateral malleolus (29); 3) tibialis anterior (TA), four fingers' breadth below the tibial tuberosity and one finger's breadth lateral to the shaft of the tibia (36); and 4) medial gastrocnemius (MG), five fingers' breadth below the popliteal crest on the medial half of the calf (36). In addition to the ankle muscles, activation in the TFL was included to investigate possible involvement of hip muscles in the control of landing. TFL is associated with hip abduction (30), and provides the largest valgus moment arm of all the muscles at the knee, which helps to stabilize the knee during the stance phase of running (28). Hence, TFL may contribute to leg-pelvis stiffness and knee joint stabilization along the frontal plane during drop landing. The surface electrodes were held in place with double adhesive tape, and raw EMG were checked for artifacts at the start of each recording session.

EMG signals were band-pass filtered at 10-500 Hz. The total gain of the system was 4048×. The filtered signals were fed into a 12-bit analog-to-digital board (DI-720P, Dataq Instruments, Inc.) at a sampling rate of 1000 Hz per channel and were stored in a portable computer for offline analysis.

Back to Top | Article Outline

Data Reduction and Analysis

Muscle response latencies.

EMG signals were full-wave rectified, then smoothed by a low-pass filter at 50 Hz. EMG response latency was defined as the time when the rectified EMG signal exceeded the mean background level plus three standard deviations (SD) for a duration of 25 ms (22). Baseline activity was determined by averaging the EMG signals for 50 ms before the drop. Response onset latencies were computed with respect to the moment of release after unexpected drops (20,33), but with respect to the instant of landing during self-initiated drops (15,38), using a customized software program written in Labview (v6.0; National Instruments, Austin, TX), which also provided graphic displays on the computer monitor for offline visual verification of response latencies. All the outcome measures described above after unexpected and self-initiated drops were averaged for each subject, except for the first four trials, to ensure consistent performance (20). Indeed, results from our pilot study, conducted in seven subjects before the main study reported here, indicate moderate to high reproducibility of EMG response latencies in the four lower-limb muscles studied, with ICC ranging from 0.60 to 0.99.

Back to Top | Article Outline

Cocontraction between TA and PER muscles, and TA and MG muscles.

Anatomically, TA acts as an ankle dorsiflexor and invertor, PER functions as an ankle plantarflexor and evertor, and MG is the main shock-absorption muscle on landing. Hence, cocontraction of TA and PER would affect ankle stiffness along both sagittal and frontal planes, whereas cocontraction of TA and MG might have greater influence on the impact force on landing. The amplitude of EMG activities was normalized with two reference values as follow:

1) The largest root mean square (RMS) EMG value of each muscle: The RMS EMG was the square root of the mean square EMG value for the middle 2-s window of a 6-s period when EMG activities reached a plateau during MVIC (5). The trial with the greatest value of the RMS EMG from the three testing trials was considered the reference measure for its corresponding muscle, termed largest RMS EMG here. Visual inspection of the EMG traces was conducted to rule out the occurrence of single spikes and to ensure the absence of other artifacts. 2) The ensemble peak EMG amplitudes obtained from unexpected drops: The prelanding period was normalized to 100% and divided into 250 equal bins. The amplitudes of the rectified EMG signals obtained from unexpected drops were averaged at the same prelanding time points to produce the ensemble EMG signals. The highest amplitude of this ensemble EMG signal was referred here as ensemble peak amplitude. The normalized TA, PER, and MG EMG were integrated over time from response onset to the moment of landing, and were compared electronically. TA/PER cocontraction index (CoI) was defined here as the ratio of twice the normalized EMG activity of the least active muscle (either TA or PER), divided by the sum of normalized EMG activities of the two muscles (TA and PER) (9). A similar approach was used in calculating the TA/MG CoI. The intraclass correlation coefficients for test-retest reliability were good, with ICC greater than 0.7 when normalized with either MVIC or ensemble peak amplitudes.

Back to Top | Article Outline

Peak vertical impact force.

Two forceplates were used to record the vertical impact force on landing from each leg. Force signal was sampled at 1000 Hz and stored for offline analysis (Dataq Instruments, Inc.). Because prelanding muscle contractions was thought to be responsible for controlling the early deceleration phase associated with landing from self-initiated drops (31), the magnitude of the first (initial) peak was used as one of the outcome measures. This measure was also found to have good reproducibility, with ICC > 0.81 in both legs, in the aforementioned pilot study that we conducted on seven subjects. To ensure valid comparison across the two subject groups, the magnitude of this initial peak force was normalized with respect to each subject's body weight (BW).

Back to Top | Article Outline

Statistical Analysis

Independent t-tests were used to compare age, height, weight, years of participation in basketball competition, and hours of practice per week between the two tested groups. Repeated-measures analysis of variance tests (ANOVA) were used to compare EMG response latencies, TA/PER, and TA/MG CoI, as well as normalized ground-reaction force on landing between the healthy and the injured groups after unexpected and self-initiated drop. Leg was regarded as a within factor, group (healthy and BMAS) as a between-group factor, and height as a covariate. The level of significance was set at P = 0.05. When a statistically significant difference was detected, post hoc independent t-tests were conducted to find out the side of the leg (left or right) for which differences existed. A significant α level to 0.025 was used for adjusting the level of significance to the comparisons between two legs.

Back to Top | Article Outline


Unexpected drops were triggered when the experimenter switched off the electrical supply to the electromagnetic suspension system. Figure 2 illustrates the traces of the right and left lower-limb muscle EMG responses from the onset of drop to the instant of landing for a healthy (solid lines) and BMAS (dotted lines) subject after an unexpected drop. After the release, there was a period of negligible EMG activities, followed by simultaneous increase in the EMG activities of all four tested muscles. Both the healthy and injured groups seemed to have similar muscle recruitment patterns in response to the unexpected drop. The impact forces on landing of these two representative subjects are also shown in Figure 2 (bottom traces). Note that the vertical ground-reaction forces increased rapidly on landing, and reached a peak at about 13-15 ms after landing.



Table 1 presents the group mean values (± SD) for the prelanding lower-limb EMG response latencies with respect to the moment of release, the TA/PER and TA/MG CoI, and the associated impact forces in both healthy and BMAS groups after unexpected drops. The EMG responses latencies in the lower-limb muscles were, on average, similar in both healthy and BMAS subjects. When comparing the two groups, no significant difference was found in the latencies of the lower-limb EMG responses to unexpected drops (using repeated-measures ANOVA, P > 0.05; Table 1). The TA/PER and TA/MG CoI of the right and the left ankles were also similar between the healthy and the MAS groups (P > 0.05, Table 1). Despite the averaged normalized peak forces being higher in the MAS than in the healthy group, by 6 and 15% in the right and left legs, respectively (cf. Table 1 where BW should be read as times body weight), post hoc tests revealed that such differences did not reach statistical significance in either ankle between groups (P > 0.025).



Figure 3 illustrates the lower-limb muscle EMG responses and the impact force on landing after self-initiated drops. Self-initiated drops were triggered when the subject himself switched off the circuit to the electromagnetic suspension. Similar to the unexpected drops, all tested muscles were relaxed and no muscle activities were recorded before release. However, unlike the unexpected drops, the TFL and the TA muscles of the two subjects seemed to respond at difference latencies. In the healthy subject (solid traces), the EMG activities of the TFL, PER, and MG muscles began at relatively similar latencies, whereas those of the TA muscle appeared later and closer to the moment of landing. In the injured subject (dotted traces), the prelanding TA EMG response was activated earlier (i.e., longer preparation for landing), whereas the TFL EMG response appeared later (e.g., closer to the moment of landing) when compared with those in the healthy subject (solid traces). In contrast, PER and MG EMG response latencies were, on average, similar between healthy and BMAS subjects. The impact forces on landing of these two representative subjects are also shown in Figure 3 (bottom traces). The first peaks in both legs seemed to be higher in the injured player when compared with those of the healthy player.



Table 2 presents the group mean values (± SD) for the prelanding lower-limb EMG response latencies, TA/PER and TA/MG CoI with respect to landing, and the associated impact forces in both study groups after self-initiated drops. Significant differences in TA (P = 0.001) and TFL (P = 0.015) response latencies were found between the two groups, using repeated-measures ANOVA. Post hoc tests revealed that TA EMG response were activated significantly earlier relative to the moment of landing, shown as 0 ms in Figure 3 (from −82.0 to −120.1 ms in the right ankles, and from −88.9 to −144.1 ms in the left ankles, P = 0.019 and 0.000 in the right and left ankles, respectively; Table 2). TFL EMG response were activated significantly later relative to the moment of landing in the left leg, and they showed a strong trend in the right leg (from −170.6 to −144.4 ms, and from −171.5 ms to −135.4 ms, for the right and left legs, respectively, P = 0.047 and 0.014; Table 2). Repeated-measures ANOVA also showed significant differences in TA/PER CoI and TA/MG CoI between the two groups (P < 0.05). Post hoc tests further revealed that the between-group differences were statistically significant in the left TA/PER CoI and TA/MG CoI (P < 0.025; Table 2), but not in the right ankle (P > 0.025, Table 2). The averaged normalized peak forces were found to be higher in the BMAS group when compared with those in the healthy group (P < 0.05, Table 2). Post hoc tests indicated that the increases were significant in the right leg but showed only a strong trend in the left leg (by 23%, P = 0.021 for the right leg; and by 8%, P = 0.031 for the left leg).



Back to Top | Article Outline


The objectives of this study were to investigate modulation in the prelanding muscle activities and their associated impact force on landing in basketball players with a history of multiple ankle sprains during unexpected and self-initiated drops. Our previous investigation indicated that prelanding muscle response latencies were modulated during self-initiated drops in healthy basketball players (15). Findings from the present study show that basketball players with multiple ankle sprains modulated their prelanding lower-limb EMG response latencies during expected (self-initiated) but not unexpected drops and that such modulations occurred in both injured and noninjured joints.

Back to Top | Article Outline

Modulation in preparatory muscle activities after unexpected and self-initiated drops.

In this study, unexpected drop was initiated by the experimenter. Even though a verbal warning signal was given before the experimenter initiated the drop, subjects could not predict the exact moment of release because of the variable time lag. The drop-evoked lower-limb muscle responses could be regarded as triggered reactions of planned muscle activities aimed to ensure a safe landing under less predictable conditions than self-initiated drops. Similar muscle response onsets and cocontraction of the ankle muscle pairs between the healthy and injured groups suggest that both groups used similar movement strategies to stiffen the legs when they could not predict the exact moment of landing. During self-initiated drops, when the injured players were anticipating the drops, modulation of some prelanding muscle response latencies and increases in CoI of the left ankle muscle pairs were observed. There are two possible explanations underlying the observed modulations: 1) a decrease in joint proprioceptive inputs, and 2) a possible increase in joint laxity of the injured (ankle) joint. As mentioned before, we previously have demonstrated the presence of ankle joint proprioceptive deficits in basketball players with multiple ankle sprains (16). In this connection, Stroeve (39) has demonstrated that reduced feedback information to a simulated musculoskeletal controller caused an increase in cocontraction of the antagonistic muscle pair of a preplanned movement. The author further suggest that such changes increase joint stiffness, and the musculoskeletal system is less vulnerable to nonoptimal control decisions based on limited feedback. Prelanding muscle activation to self-initiated drops is preplanned (31,38); the proprioceptive deficits associated with ankle sprains in the injured basketball players could be one of the factors causing them to adopt a cocontraction strategy in preparation for landing from self-initiated drops. In addition, joint laxity has been detected in one third of the subjects with functionally unstable ankles (27). The increase in the cocontraction of the left ankle may be a subconscious effort to protect the injured joint from excessive movement caused by increased joint laxity. In this connection, Beiser et al. (3) have demonstrated that healthy subjects can make postural adjustments to minimize the forthcoming perturbation during anticipated running and cutting movements. Results from this study suggest that cocontraction of the ankle muscles could be one component of the landing strategy during self-initiated drops for subjects who have suffered bilateral multiple ankle sprains, to minimize excessive movement at the ankle joint.

In addition to the injured joint, our results further demonstrate that modulation of the prelanding muscle responses involved the injured ankle and the uninjured hip joints. A trend of delayed onset in the right leg, with a significant delay in the left leg relative to the moment of landing in TFL EMG response latencies, was observed in the injured group compared with the healthy group (P = 0.047 and 0.014, respectively; Table 2). In this connection, trunk movement such as leaning forward or backward before platform translation was associated with an increase in hip muscle activation and delayed and suppressed ankle muscle responses (35). Furthermore, standing with a narrow base increased the EMG magnitudes, particularly those of proximal leg muscles (21). A significant delay in the left prelanding TFL response in the injured group might be attributable to alterations in trunk orientation or/and changes in the width of landing legs on landing. Kinematic data from the trunk and lower legs during landing were not investigated in the present study. Further investigation is suggested to include motion analysis using high-speed cameras to reveal possible relationships among trunk orientation, stance width, and muscle activation for comparison between healthy and injured players.

Our findings also show that the magnitude of the normalized vertical peak force was significantly greater in the injured group compared with that of the healthy group during landing from self-initiated drops (P < 0.05). Post hoc tests indicate that the changes in the right leg reached a significant level (P = 0.012, Table 2) and the left leg showed a strong trend of increase (P = 0.031, Table 2). Prelanding muscular responses were thought to enhance joint stability on landing and to prepare the body to absorb the impact force associated with landing (31). The increase in the impact force on landing might partly be attributable to the increase in cocontraction of the ankle dorsiflexor and plantarflexor in the injured players when compared with that of the healthy players. Indeed, we have recently shown that a positive relationship exists between prelanding cocontraction of ankle dorsiflexor and plantarflexor and impact force on landing from self-initiated drops (17).

Back to Top | Article Outline

Other considerations.

In the present study, basketball players with bilateral ankle sprains were recruited and were compared with age- and activity-matched players. Most of the previous studies on platform perturbation have been performed with subjects who had sustained unilateral ankle sprains. The uninjured contralateral ankle was normally used as a control, to avoid intersubject confounding factors such as age and activity level. However, Leanderson et al. (26) found that the majority (of 76%) of team basketball players had sustained bilateral, rather than unilateral, ankle sprains. A similar observation was noted in our players. Hence, lower-limb muscle responses of the healthy players similar in age, gender, and activity level served as reference for the injured players in our study.

The analysis of prelanding neuromuscular strategies between healthy and injured players was based on the EMG signals captured from surface electrodes. Note that drop landing is a dynamic task, and changes in joint angle during the movement could cause electrode shift (12) and modulations in the conductivity of tissues between surface electrodes and muscle fibers (34). Electrode shift was minimized with the use of double adhesive tape to hold the surface electrodes in place. However, unwanted artifact could still occur because of changes in the geometric relationship between electrodes and muscle fibers (12). In addition, timing of the muscle response could be influenced by amplitude cancellation, which is influenced by a change in tissue geometry (11). Nevertheless, differences in TA and TFL muscle response onsets between healthy and injured groups were more than 25 ms, which is higher than the resolution in using EMG to identify muscle response onset (about 10 ms) (10). Normalization of EMG amplitude is essential to eliminate the physiological factors that influence its amplitude. During a dynamic task, normalization using reference values other than the amplitude obtained during MVIC has been proposed by Yang and Winter (41). In our study, normalization of EMG responses using both MVIC and ensemble peak amplitude recorded during unexpected drops was computed for cross-reference. Differences in the cocontraction of the muscle pairs were observed between the two normalization methods. However, both methods produced significant differences in the CoI in the left but not the right ankle between the two study groups. Aside from normalization, amplitude cancellation (11), volume-conduction properties (34), and electrode location could influence the amplitude of the signal recorded during dynamic contraction (10). New approaches for detection and interpretation of dynamic EMG, such as the use of high-density surface EMG and wavelet analysis, are being explored (10). Such novel methods could further increase the validity of signal detection and processing during dynamic tasks.

Back to Top | Article Outline

Clinical implications.

Both healthy and injured subjects for this study participated in basketball matches at a competitive level, and they practiced basketball for an average of 5 h·wk−1. In other words, injured athletes were still able to perform high-level activities such as landing from jumps, albeit with a somewhat different strategy when landing from self-initiated ones. The present findings suggest that the landing strategy adopted by the injured basketball players could be a "feed-forward" strategy in anticipation of the impact on landing. This may protect the joint from excessive joint motion, because the increase in the cocontraction between the ankle muscle pairs could take over the function of damaged ligaments to maintain joint stability. However, such an adopted strategy is less efficient (25) in that greater activation of the antagonist muscle is required. In addition, the increased stiffness would reduce the compliance of ankle joints, making the landing foot less adaptable to uneven surfaces. The increase in the impact force on landing with long-term basketball practice could further lead to earlier degenerative changes in the affected joints (37). These findings suggested that a comprehensive rehabilitation program for basketball players with bilateral ankle sprains should promote a proper balance between ankle stability and joint compliance at the same time as minimizing the impact force on landing. More specifically, the program should incorporate exercises to enhance the coordination between lower-limb muscles in a task-specific manner for landing, in addition to training of ankle joint proprioception as suggested by our previous studies (16,17).

In conclusion, modulations of prelanding EMG responses were observed in both injured (ankle) and noninjured (hip) joints during self-initiated but not unexpected drops in basketball players with bilateral ankle sprains when compared with the healthy players. After self-initiated drops, the TA response was activated earlier and the left TFL closer to the moment of landing; the prelanding CoI between the left TA and PER, and the left TA and MG, were significantly greater in the injured than in the healthy players. In addition, the impact force on landing was found to be greater in the injured basketball players.

The authors thank the Hong Kong Polytechnic University for financial support of this study through an Area of Strategic Development Grant to C.W.Y. Hui-Chan and team. The authors also thank the basketball players for their participation.

No commercial party having a direct financial interest in the research findings reported here has or will confer a benefit upon the authors or upon any organization with which the authors are associated.

Back to Top | Article Outline


1. Ashton-Miller, J. A., R. A. Ottaviani, C. Hutchinson, and E. M. Wojtys. What best protects the inverted weightbearing ankle against further inversion? Evertor muscle strength compares favourably with shoe height, athletic tape, and three orthoses. Am. J. Sports Med. 24:800-809, 1996.
2. Baldwin, F. C., and J. Tetzlaff. Historical perspectives on injuries of the ligaments of the ankle. Clin. Sport Med. 1:3-12, 1982.
3. Besier, T. F., D. G. Lloyd, T. R. Ackland, and J. L. Cochrane. Anticipatory effects on knee joint loading during running and cutting maneuvers. Med. Sci. Sports Exerc. 33:1176-1181, 2001.
4. Braun, B. L. Effects of ankle sprain in a general clinic population 6 to 18 months after medical evaluation. Arch. Fam. Med. 8:143-148, 1999.
5. Cram, J. R., G. S. Kasman, and J. Holtz. Instrumentation and atlas for electrode placement. In: Introduction to Surface Electromyography, J. R. Cram (Ed.). Gaithersburg, MD: Aspen, 1998, pp. 43-80, 371-375.
6. Dietz, V., W. Zijlstra, T. Prokop, and W. Berger. Leg muscle activation during gait in Parkinson's disease: adaptation and interlimb coordination. Electroencephalogr. Clin. Neurophysiol. 97:408-415, 1995.
7. Ebig, M., S. M. Lephart, R. G. Burdett, M. C. Miller, and D. M. Pincivero. The effect of sudden inversion stress in EMG activity of the peroneal and tibialis anterior muscles in the chronically unstable ankle. J. Orthop. Sports Phys. Ther. 26:73-77, 1997.
8. Enoka, R. M., and A. J Fuglevand. Neuromuscular basis of the maximum voluntary force capacity of muscle. In: Current Issues in Biomechanics, M. D. Grabiner (Ed.). Champaign, IL: Human Kinetics, 1993, pp. 215-235.
9. Falconer, K., and D. A. Winter. Quantitative assessment of co-contraction at the ankle joint in walking. Electroencephalogr. Clin. Neurophysiol. 25:135-149, 1985.
10. Farina, D. Interpretation of the surface electromyogram in dynamic contractions. Exerc. Sport Sci. Rev. 34:121-127, 2006.
11. Farina, D., R. Merletti, and R. M. Enoka. The extraction of neural strategies from the surface EMG. J. Appl. Physiol. 96:1486-1495, 2004.
12. Farina, D., R. Merletti, M. Nazzaro, and I. Caruso. Effect of joint angle on EMG variables in leg and thigh muscles. IEEE Eng. Med. Biol. Mag. 20:62-71, 2001.
13. Fernandes, N., G. T. Allison, and D. Hopper. Peroneal latency in normal and injured ankles at varying angles of perturbation. Clin. Orthop. Relat. Res. 375:193-201, 2000.
14. Freeman, M. A. R., M. R. E. Dean, and I. W. F. Hanham. The etiology and prevention of functional instability of the foot. J. Bone Joint Surg. 47B:678-685, 1965.
15. Fu, S. N., and C. W. Y. Hui-Chan. Mental set can modulate response onset in the lower limb muscles to falls in humans. Neurosci. Lett. 321:77-80, 2002.
16. Fu, A. S. N., and C. W. Y. Hui-Chan. Ankle joint proprioception and postural control in basketball players with bilateral ankle sprains. Am. J. Sport Med. 33:1174-1182, 2005.
17. Fu, S. N., and C. W. Y. Hui-Chan. Are there any relationships among ankle proprioception acuity, pre-landing ankle muscle responses, and landing impact in man? Neurosci. Lett. 417:123-127, 2007.
18. Gerber, J. P., G. N. Williams, C. R. Scoville, R. A. Arciero, and D. C. Taylor. Persistent disability associated with ankle sprains: a prospective examination of an athletic population. Foot Ankle Int. 19:653-660, 1998.
19. Glencross, D., and E. Thornton. Position sense following joint injury. J. Sports Med. Phys. Fitness 21:23-27, 1981.
20. Greenwood, R., and A. Hopkins. Muscle responses during sudden falls in man. J. Physiol. 254:507-518, 1976.
21. Henry, S., J. Fung, and B. F. B. Horak. Effects of stance width on multidirectional postural responses. J. Neurophysiol. 85:559-570, 2001.
22. Hodges, P. W., and B. H. Bui. A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography. Electroencephalogr. Clin. Neurophysiol. 101:511-519, 1996.
23. Jerosch, J., I. Hoffstetter, H. Bork, and M. Bischof. The influence of orthoses on the proprioception of the ankle joint. Knee Surg. Sports Traumatol. Arthrosc. 3:39-46, 1995.
24. Konradsen, L., and J. B. Ravn. Prolonged peroneal reaction time in ankle instability. Int. J. Sports Med. 12:290-292, 1991.
25. Konradsen, L., G. Peura, B. Beynnon, and P. Renstrom. Ankle eversion torque response to sudden ankle inversion torque response in unbraced, braced, and pre-activated situations. J.Orthop. Res. 23:315-321, 2005.
26. Leanderson, J., G. Nemeth, and E. Eriksson. Ankle injuries in basketball players. Knee Surg. Sports Traumatol. Arthrosc. 1:200-202, 1993.
27. Lentell, G., B. Baas, D. Lopez, L. McGuire, M. Sarrels, and P.Snyder. The contributions of proprioceptive deficits, muscle function, and anatomic laxity to functional instability of the ankle. J. Orthop. Sports Phys. Ther. 21:206-215, 1995.
28. Lloyd, D. G., and T. S. Buchanan. Strategies of muscular support of varus and valgus isometric loads at the human knee. J.Biomech. 34:1257-1267, 2001.
29. Lynch, S. A., U. Eklund, D. Gottlieb, and A. F. H. Renstrom. Electromyographic latency changes in the ankle musculature during inversion moments. Am. J. Sports Med. 24:362-369, 1996.
30. McClay, I. S., K. J. Lake, and P. R. Cavanagh. Muscle activity in running. In: Biomechanics in Distance Running, P. R. Cavanagh (Ed.). Champaign, IL: Human Kinetics Book, 1990, pp. 165-186.
31. Mckinley, P., and A. Pedotti. Motor strategies in landing from a jump: the role of skill in task execution. Exp. Brain Res. 90:427-440, 1992.
32. McNitt-Gray, J. L., D. M. Hester, W. Mathiyakom, and B. A. Munkasy. Mechanical demand and multijoint control during landing depend on orientation of the body segments relative to the reaction force. J. Biomech. 34:1471-1482, 2001.
33. Melvill-Jones, G., and G. D. Watt. Muscular control of landing from unexpected falls in man. J. Physiol. 219:729-737, 1971.
34. Mesin, L., M. Joubert, T. Hanekon, R. Merletti, and D. Farina. A finite element model for describing the effect of muscle shortening on surface EMG. IEEE Trans. Biomed. Eng. 53:593-6000, 2006.
35. Moore, S. P., F. B. Horak, and L. M. Nashner. Influence of initial stance position on human postural responses. Abstr. Soc. Neurosci. 12:1301, 1986.
36. Perotto, A. O. Leg. In: Anatomical Guide for the Electromyographer: The Limbs and Trunk, A. O. Perotto (Ed.). Springfield, MA: Charles C. Thomas, 2005, pp. 177-178, 191-193.
37. Radin, E. L. Nature of mechanical factors causing degeneration of joints in the hip. In: Proceedings of the 2nd Open Scientific Meeting of the Hip Society. St. Louis, MO: Mosby, 1974, pp.76-81.
38. Santello, M., and M. J. Mcdonagh. The control of timing and amplitude of EMG activity in landing movements in humans. Exp.Physiol. 83:857-874, 1998.
39. Stroeve, S. Learning combined feedback and feedforward control of a musculoskeletal system. Biol. Cybern. 75:73-83, 1996.
40. Winter, D. A., A. E. Patla, F. Prince, M. Ishac, and K. Gielo-Perczak. Stiffness control of balance in quiet standing. J.Neurophysiol. 80:1211-1221, 1998.
41. Yang, J. F., and D. A. Winter. Electromyographic amplitude normalization methods: improving their sensitivity as diagnostic tools in gait analysis. Arch. Phys. Med. Rehabil. 65:517-521, 1984.


©2007The American College of Sports Medicine