Mono- and Biarticular Muscle Activity During Jumping in Different Directions

Transcription

1 Muscle Activity During Jumping 205 JOURNAL OF APPLIED BIOMECHANICS, 2003, 19, by Human Kinetics Publishers, Inc. Mono- and Biarticular Muscle Activity During Jumping in Different Directions Stephanie L. Jones and Graham E. Caldwell University of Massachusetts Amherst This study examined the role of mono- and biarticular muscles in control of countermovement jumps (CMJ) in different directions. It was hypothesized that monoarticular muscles would demonstrate the same activity regardless of jump direction, based on previous studies which suggest their role is to generate energy to maximize center-of-mass (CM) velocity. In contrast, biarticular activity patterns were expected to change to control the direction of the ground reaction force (GRF) and CM velocity vectors. Twelve participants performed maximal CMJs in four directions: vertical, forward, intermediate forward, and backward. Electromyographical data from 4 monoarticular and 3 biarticular lower extremity muscles were analyzed with respect to segmental kinematics and kinetics during the jumps. The biarticular rectus femoris (RF), hamstrings (HA), and gastrocnemius all exhibited changes in activity magnitude and pattern as a function of jump angle. In particular, HA and RF demonstrated reciprocal trends, with HA activity increasing as jump angle changed from backward to forward, while RF activity was reduced in the forward jump condition. The vastus lateralis and gluteus maximus both demonstrated changes in activity patterns, although the former was the only monoarticular muscle to change activity level with jump direction. Mono- and biarticular muscle activities therefore did not fit with their hypothesized roles. CM and segmental kinematics suggest that jump direction was initiated early in the countermovement, and that in each jump direction the propulsion phase began from a different position with unique angular and linear momentum. Issues that dictated the muscle activity patterns in each jump direction were the early initiation of appropriate forward momentum, the transition from countermovement to propulsion, the control of individual segment rotations, the control of GRF location and direction, and the influence of the subsequent landing. Key Words: muscle coordination, muscular control, electromyography, stretch shorten cycle The authors are with the Dept. of Exercise Science, Totman Bldg., University of Massachusetts Amherst, Amherst, MA

2 206 Jones and Caldwell Introduction In multisegment movements, the contributions of monoarticular muscles (those crossing one joint) may differ from those of biarticular muscles (van Ingen Schenau, 1989; van Ingen Schenau, Bobbert, & Rozendal, 1987; van Ingen Schenau, Dorssers, Welter, et al., 1995). For example, it has been suggested that biarticular muscles transfer energy generated by monoarticular extensors to the adjacent distal joint to produce powerful lower extremity extension during vertical jumping (Bobbert & Van Ingen Schenau, 1988). This biarticular energy transfer supposedly results in more effective lower extremity extension and greater jump heights (van Soest, Schwab, Bobbert, & van Ingen Schenau, 1993). Another benefit is that biarticular muscles may provide a means to adjust or tune adjacent joint moments so that external forces can be applied in a specific direction (Jacobs & van Ingen Schenau, 1992). For example, consider the task of exerting forces in different directions against the ground with the foot while standing upright. A forward-directed force vector would require both hip flexor and knee extensor moments, a combination suited perfectly for the biarticular rectus femoris. Likewise, the biarticular hamstrings muscles could contribute to both hip extensor and knee flexor moments needed for a force vector applied in the posterior direction. A question that has not been addressed is how monoarticular and biarticular muscles are coordinated when there are possible conflicts in their supposed roles. In vertical jumping, the timely activation of the biarticular gastrocnemius transfers energy generated by the monoarticular knee extensors to help ankle plantarflexion late in propulsion (Bobbert & Van Ingen Schenau, 1988; van Ingen Schenau, 1989; van Ingen Schenau et al., 1987). In this case the ground reaction force (GRF) vector is directed vertically to accelerate the body center of mass (CM) upward. Jumping in other directions, such as in a standing long jump, still requires powerful lower extremity extension to propel the body, but may need differing combinations of joint moments to properly direct the GRF and CM velocity vectors. For example, Ridderikhoff, Batelaan, and Bobbert (1999) reported distinct changes in hip, knee, and ankle moments as jumpers altered the direction of the squat jump from vertical to forward, and they attributed these adaptations to modification of GRF direction. Therefore a directional jump may present conflicting requirements for biarticular muscles, juxtaposing energy transfer and directional control mechanisms. This is similar to the conflict between energy generation and pedal force directional control which may be responsible for antagonistic co-contraction between monoarticular and biarticular muscles observed in cycling (van Ingen Schenau, Boots, de Groot, Snackers, & Van Woensel, 1992; van Ingen Schenau et al., 1995). How does biarticular muscle usage change with an alteration in jump direction? Does the need for directional control modify the role of biarticular muscles? Does the role of monoarticular muscles change with jump direction? Although previous studies of jumping at modified takeoff angles have provided insight into the movement strategies employed in directional jumps (Jensen & Phillips, 1991; Ridderikhoff et al., 1999; Selbie & Caldwell, 1998), we are aware of none that have examined the coordination of monoarticular and biarticular muscles in this context. Therefore, the aim of the present study was to examine the activation patterns of mono- and biarticular muscles during countermovement jumps with different takeoff angles. Based on previous research, we hypothesized that mono-

3 Muscle Activity During Jumping 207 articular muscles would be used in a similar manner to provide maximal energy generation for lower extremity extension, regardless of jump direction (Jacobs & van Ingen Schenau, 1992; van Ingen Schenau, 1989; van Ingen Schenau et al., 1992; 1995). In contrast, we expected that biarticular activation patterns would vary as the change in takeoff angle altered the GRF directional constraint on energy transport across adjacent joints. Methods Twelve young adults (11 M, 1 F) drawn from the local university community (mean age 26 ± 5 years; mass 78 ± 9 kg; height 182 ± 8 cm) participated in this study. Each of them participated recreationally in sports that involved jumping, including basketball and volleyball. They all provided written informed consent in accordance with university policy. For habituation, the participants were required to complete 5 to 10 practice jumps in each of four directions. They performed maximal-effort countermovement jumps at a specified angle from the horizontal with their arms clasped behind their back, beginning from a neutral standing position on a force platform (Advanced Medical Technologies Inc., Watertown, MA). The jump conditions were designated as vertical (VJ: 90, 1.57 rad), forward (FJ: 45, 0.79 rad), intermediate forward (IJ: 70, 1.22 rad), and backward (BJ: 120, 2.09 rad). The participants were instructed to jump as high as possible for the vertical jump, and for maximal horizontal displacement for the forward and backward jumps. For the intermediate forward jump, they were instructed to jump as high as possible in the direction of a target on the ceiling. They were given immediate achieved-angle feedback using a computer program that integrated the GRF record to determine CM velocity vector direction. Participants were considered habituated when they could perform each jump repeatedly within 5 of the target angle. During a second experimental session, reflective markers were placed on anatomical landmarks (acromion process, greater trochanter, lateral femoral condyle, lateral malleolus, lateral aspect of the talus, and 5th metatarsal head) to delineate the head/arms/trunk (HAT), thigh, leg, and foot segments in the sagittal view. Preamplified electrodes (Therapeutics Unlimited Inc., Iowa City, IA) were applied to four monoarticular muscles: gluteus maximus (GM)*, vastus lateralis (VL), soleus (SO), and tibialis anterior (TA), and to three biarticular muscles: rectus femoris (RF), medial hamstrings (HA), and lateral head of gastrocnemius (GA), using standard electrode placement and preparation (Winter, 1990). The electromyographical (EMG) signals were further amplified (CMRR: 87 db at 60 Hz, input impedance: >25 MΩ, bandwidth: Hz; Therapeutics Unlimited) with gains set to optimize signal resolution. The participants were required to perform 3 successful jumps in each direction. Successful jumps were those in which (a) takeoff occurred within the acceptable target range (±5 ); (b) the participant and experimenters agreed that the effort was close to maximal; and (c) the participant landed in a controlled and stable manner. Kinematic data were recorded using a 200-Hz digital camera (Qualisys *GM data were later found to be corrupt for 5 participants, and therefore are reported for the other 7 participants only.

4 208 Jones and Caldwell Inc., Glastonbury, CT) interfaced with an analog-to-digital converter that collected force and EMG signals at 1,000 Hz in synchrony with the kinematic data. The order of the conditions was balanced across participants. Marker position and segment angle data were filtered using a dual pass 2nd-order Butterworth low-pass filter with a cutoff frequency of 15 Hz. A foursegment, rigid link model was constructed for each jumper, with segment centers of mass determined using regression equations from the literature (Dempster, 1955; Winter, 1990). First and second time derivatives of the linear and angular position data were calculated numerically with a central-difference technique to provide velocity and acceleration data. Force data were low-pass filtered at 10 Hz (2ndorder Butterworth) and resampled at 200 Hz in synchrony with the kinematic data. All filter cutoff frequencies were chosen based on frequency content (FFT techniques) and residual analysis (Winter, 1990). Segment kinematics and force data were used to compute hip, knee, and ankle moments through inverse dynamics (Bresler & Frankel, 1950). For each jump, countermovement duration was defined from movement initiation (deviation of the vertical CM velocity from zero) until the lowest CM position (when vertical CM velocity 0). Propulsion duration spanned the remaining time until takeoff (identified when the unfiltered vertical GRF fell below 10 N). Linear forward momentum was calculated as the product of the jumper s mass and CM horizontal velocity. Joint moments were integrated from jump initiation to takeoff to calculate angular impulse at each joint and were subsequently separated into three phases: early countermovement (first 50%), late countermovement (second 50%), and propulsion. EMG signals were rectified and smoothed at 10 Hz in order to create linear envelopes, and were standardized to the maximum value for that muscle from the vertical trials for each jumper. For qualitative analysis, ensemble kinematic, kinetic, and EMG curves were calculated for each condition. EMG parameters were calculated from EMG onset to takeoff for all muscles. EMG onset was determined visually for all three trials as the point at which the linear envelope deviated sharply from the baseline level. Repeated onset assessment showed this to be a superior method compared to objective threshold criteria that tended to produce false onset determinations. Measures of peak and average EMG magnitude were compared using a twofactor (takeoff angle, trial) repeated-measures ANOVA model. Timing measures for EMG onsets and time to peak in relation to takeoff were compared using a three-factor (muscle, takeoff angle, trial) repeated-measures ANOVA. Similar comparisons were performed on kinematic and joint kinetic measures to help interpret the EMG results. Significant results, p < 0.05, were tested using Tukey s HSD test, or individual t-tests when the sphericity assumption was violated. Results The GRF and CM velocity vectors at takeoff confirmed that the jumpers were successful in achieving each target direction within a 4 (0.07 rad) range (Figure 1 and Table 1). The similarity in CM takeoff speed near 2.75 m/s (p 0.15), and GRF peak magnitude near 1,700 N (p 0.08), suggests that the jumpers achieved consistent effort in all jump directions. Each direction produced unique CM mo-

5 Muscle Activity During Jumping 209 Figure 1 Ensemble joint coordinates in each jump condition, plotted at 100-ms intervals from countermovement initiation until takeoff. For the vertical, intermediate, and forward conditions, the penultimate and final stick figures are separated by 50 ms for better resolution. CM velocity and GRF vectors are plotted originating from the body CM and the COP, respectively. Vector magnitudes are scaled to maximum value in each condition. Horizontal bars represent average duration of countermovement (light gray) and propulsion (dark gray) phases for the 12 jumpers. The thinner intermediate gray bar indicates the range of differences in phase durations between jump conditions.

6 210 Jones and Caldwell Table 1 Mean (± SD) Values for Descriptive Measures of 4 Jumping Conditions BJ VJ IJ FJ p-value CM velocity angle * at takeoff (rad) ±0.05 a ± 0.03 b ±0.06 c ±0.07 d CM takeoff speed (m/s) ±0.27 ±0.26 ±0.25 ±0.20 Jump duration * (ms) ±140 c ±150 b ±150 b ±120 a Countermovement * duration (ms) ±120 c ±130 c ±140 b ±120 a Propulsion duration * (ms) ±44 b ±50 a ±40 a ±40 a Angle of GRF vector * at takeoff (rad) ±0.04 a ±0.03 b ±0.05 c ±0.05 d Peak magnitude of GRF (N) ±235 ±229 ±213 ±226 CM forward momentum * at end of counter- ±7.7 d ±4.9 c ±8.7 b ±13.4 a movement (kg m/s) Note: The p-values indicate statistical significance of main effect of jump direction. Superscripts indicate conditions that were significantly different (i.e., a b c d at p < 0.05). CM speed and angle were determined by differentiating the CM position record. tion trajectories (Figure 2) and timing of CM horizontal motion (Figure 3). For each jump, the CM began just behind the 5th metatarsal marker (slightly negative), with horizontal translation starting early in the countermovement. The CM horizontal position could differentiate between the four jump directions 600 ms before takeoff, 300 ms before propulsion began. At the end of the countermovement, each jump direction exhibited unique CM forward momentum appropriate for the eventual take-off direction (Table 1: BJ = 15.0 kg m/s; VJ = 14.3; IJ = 29.3; FJ = 55.9). Several timing parameters varied with jump direction. For example, the backward jump duration was shorter (862 ms) than the other jumps, caused by reductions in both propulsion and countermovement duration, while the forward jump was longer (1,000 ms), due to an extended (70- to 100-ms longer) countermovement phase (Table 1). The TA was activated near the initiation of countermovement and was the first muscle to reach peak activity (Table 2 and Figure 4). Jumpers also produced a small early RF activity burst, although it is not evident in the ensemble patterns because its timing varied between individuals. By late in the countermovement, all muscles were active to decelerate the downward motion of the CM and initiate

7 Muscle Activity During Jumping 211 Figure 2 Ensemble body CM X and Y coordinates (m). Backward jump (thin solid line), vertical jump (thick solid line), intermediate jump (crosses), and forward jump (open circles) conditions are plotted from quiet stance to takeoff. Figure 3 Ensemble time histories of horizontal body CM position (m) relative to the horizontal position of the 5th metatarsal head (5th Met). Backward jump (thin solid line), vertical jump (thick solid line), intermediate jump (crosses), and forward jump (open circles) conditions are plotted from quiet stance to takeoff. Horizontal bars represent average duration of countermovement (light gray) and propulsion (dark gray) phases for the 12 jumpers. Positive position values indicate that body CM is in front of the fifth metatarsal head; negative values indicate that body CM is behind.

8 212 Jones and Caldwell Table 2 Mean (± SD) Values for Time to Peak Linear Envelope EMG and for Average and Peak EMG Magnitude as Percent of Maximum Vertical Trial Value per Jumper Muscle BJ VJ IJ FJ p-values Time to peak EMG (ms) relative to takeoff Two-joint RF 170 ± ± ± ± HA 154 ± ± ± ± GA 126 ± ± ± ± One-joint GM 155 ± ± ± ± VL 187 ± 133 b 209 ± 139 a,b 225 ± 150 a 230 ± 145 a 0.05* SO 149 ± ± ± ± TA* 398 ± 240 b 480 ± 219 a,b 536 ± 206 a 578 ± 193 a 0.01* Average (%VJ max) Two-joint RF 30 ± 9 a 29 ± 6 a 26 ± 8 a 19 ± 5 b 0.01* HA 24 ± 9 d 30 ± 10 c 38 ± 14 b 47 ± 21 a 0.01* GA 24 ± 6 c 27 ± 8 b,c 30 ± 10 a,b 33 ± 9 a 0.01* One-joint GM 26 ± ± 8 28 ± ± VL 33 ± 11 a,b 38 ± 13 a 35 ± 13 a,b 30 ± 12 b 0.01* SO 32 ± 9 32 ± 7 34 ± 8 30 ± TA 24 ± ± 5 28 ± 9 27 ± Peak (%VJ max) Two-joint RF 93 ± 17 a 90 ± 11 a 89 ± 17 a 72 ± 18 b 0.01* HA 80 ± 37 b 86 ± 15 b 119 ± 31 a 136 ± 45 a 0.01* GA 85 ± 15 b 89 ± 11 a,b 98 ± 16 a 97 ± 14 a 0.01* One-joint GM 78 ± ± ± ± VL 85 ± ± ± ± SO 92 ± ± ± ± TA 67 ± ± ± ± *Time to peak TA activity was significantly earlier than all other muscles, p < Superscripts indicate conditions that were significantly different (i.e., a b c d at p < 0.05).

9 Muscle Activity During Jumping 213 Figure 4 Ensemble time histories of smoothed, rectified EMG curves of monoarticular GM, VL, SO, and TA muscles, and biarticular RF, HA, and GA muscles. EMG is plotted as a percentage of single trial maximum of each jumper s vertical jump. Backward jump (thin solid line), vertical jump (thick solid line), intermediate jump (crosses), and forward jump (open circles) conditions are plotted from quiet stance to takeoff. Horizontal bars represent average duration of countermovement (light gray) and propulsion (dark gray) phases for the 12 jumpers. Thin horizontal boxes represent range of onset times for each muscle across all conditions.

10 214 Jones and Caldwell subsequent propulsion. Although the extensor activity patterns displayed a general proximal-to-distal sequence, the timings of their activity peaks were statistically similar (Table 2). EMG patterns and magnitudes showed direction-related statistical differences for all three biarticular muscles, particularly HA (Figure 4 and Table 2). As takeoff direction moved from FJ to BJ, both peak and average activity increased for RF but decreased for HA. Rectus femoris activity was significantly lower in FJ than in the other jump directions as measured by both peak (72 ± 18% VJ max) and average EMG (19 ± 5% VJ max). For HA, the average EMG values decreased sequentially from FJ (47 ± 21%) to IJ (38 ± 14%) to VJ (30 ± 10%) to BJ (24 ± 9%). The peak HA values followed a similar trend, with FJ (136 ± 45%) and IJ (119 ± 31%) significantly higher compared to VJ (86 ± 15%) or BJ (80 ± 37%). The biarticular GA displayed smaller differences, with peak activity in BJ (85 ± 15%) significantly less than in IJ (98 ± 16%) and FJ (97 ± 14%). Average EMG showed a similar pattern for GA, with BJ (24 ± 6%) and VJ (27 ± 8%) significantly less than IJ (30 ± 10%) and FJ (33 ± 9%). For the biarticular muscles there were no significant differences in timing of onset or peak activity across jumping conditions. In contrast to these two-joint muscles, the monoarticular extensors GM, VL, and SO demonstrated similar peak EMG activity in all jump directions (Table 2). However, GM and VL activity patterns changed with jump direction during late propulsion (Figure 4). The VL pattern exhibited a reduction in activity after the peak that was most pronounced in FJ. The average EMG data for VL reflected these patterning differences, with the VJ (38 ± 13%) value being significantly higher than in FJ (30 ± 12%). Pattern differences were also evident in VL time to peak EMG, with the BJ condition demonstrating a later peak ( 187 ± 133 ms) than either IJ ( 225 ± 150 ms) or FJ ( 230 ± 145 ms). The average GM values were statistically similar and thus did not reflect the pattern variations seen in Figure 4. Although the mono-articular dorsiflexor TA patterns were similar, the more forward-directed conditions (FJ, IJ) demonstrated higher activity at the beginning of the countermovement, and displayed TA peak EMG values earlier (IJ: 536 ± 206 ms; FJ: 578 ± 193 ms) than in BJ ( 398 ± 240 ms). Peak and average tibialis anterior EMG data were not statistically different even though BJ appears to have had less activity during the countermovement (Figure 4). These alterations in muscle activity resulted in segmental kinematic patterns that varied with jump direction (Figures 1 and 5). Statistical differences were found between jumping conditions for peak flexion angle magnitude and timing, and angle magnitude at takeoff (Table 3). Biarticular HA and RF activity differences may be related to the need to arrest and reverse varying amounts of HAT forward flexion. The most flexion was observed in FJ (peak 0.18 ± 0.18 rad), in which HA activity was highest and RF was lowest, while the least HAT flexion was in the BJ condition (peak 0.58 ± 0.14 rad) in which HA activity was reduced. Lower extremity kinematic patterns also varied with jump direction, but unlike the HAT changes, these differences became greater as the jumps progressed. This resulted in distinctly different segment positions at the bottom of the countermovement and upon takeoff for each jumping direction. The peak thigh and leg angles displayed progressively more forward rotation as the jumping direction went from backward

11 Muscle Activity During Jumping 215 Figure 5 Ensemble time histories of HAT, thigh, leg, and foot segment angles (rad). Backward jump (thin solid line), vertical jump (thick solid line), intermediate jump (crosses), and forward jump (open circles) conditions are plotted from quiet stance to takeoff. Segment angles are measured with respect to right horizontal (inset figure). Larger values indicate more backward rotation. Horizontal bars represent average duration of countermovement (light gray) and propulsion (dark gray) phases for the 12 jumpers.

12 216 Jones and Caldwell Table 3 Mean (± SD) Values for Peak and Takeoff Segmental Flexion Angles and Timing of Peaks With Respect to Takeoff Segment BJ VJ IJ FJ p-value Peak flexion angle (rad) HAT 0.58 ± 0.14 a 0.50 ± 0.16 a 0.39 ± 0.16 b 0.18 ± 0.18 c 0.01* Thigh 2.72 ± 0.08 a 2.56 ± 0.11 b 2.44 ± 0.08 c 2.31 ± 0.07 d 0.01* Leg 1.01 ± 0.05 a 0.85 ± 0.06 b 0.76 ± 0.07 c 0.55 ± 0.07 d 0.01* Time to peak (ms) relative to takeoff HAT 327 ± ± ± ± Thigh 233 ± 32 d 295 ± 38 c 320 ± 43 b 341 ± 39 a 0.01* Leg 277 ± 50 a 208 ± 35 b 168 ± 28 c 127 ± 24 d 0.01* Flexion angle at takeoff (rad) HAT 1.17 ± 0.19 b 1.30 ± 0.10 a 1.31 ± 0.09 a 1.20 ± 0.09 b 0.01* Thigh 2.08 ± 0.07 a 1.63 ± 0.03 b 1.41 ± 0.06 c 1.18 ± 0.10 d 0.01* Leg 1.77 ± 0.06 a 1.47 ± 0.05 b 1.25 ± 0.05 c 0.85 ± 0.09 d 0.01* Foot 2.24 ± 0.10 a 1.94 ± 0.08 b 1.71 ± 0.09 c 1.34 ± 0.10 d 0.01* Note: The p-values indicate statistical significance of main effect of jump condition. Superscripts indicate conditions that were significantly different (i.e., a b c d at p < 0.05). (BJ: thigh 2.72 ± 0.08 rad; leg 1.01 ± 0.05 rad) to forward (FJ: thigh 2.31 ± 0.07 rad; leg 0.55 ± 0.07 rad). Also, peak thigh angles occurred progressively earlier as the jumping direction went from backward (BJ: 233 ± 32 ms before takeoff) to forward (FJ: 341 ± 39 ms). In contrast, the timing of peak leg angles revealed the opposite trend (BJ: 277 ± 50 ms before takeoff; FJ: 127 ± 24 ms). Overall, these segmental patterns and timings reflect unique segmental coordination for each target direction. The early RF burst and TA activity initiated the countermovement by inducing a small hip flexor moment coupled with a decrease in ankle plantarflexor moment 700 to 1,100 ms before takeoff (Figure 6). The decrease in plantarflexor moment was greatest in FJ, which demonstrated the highest early TA activity and significantly less early countermovement angular impulse than in BJ (Figure 4 and Table 4). At the same time, knee extensor moment was highest for FJ and lowest for BJ. Later in the countermovement, the knee moment differences between conditions disappeared 650 ms before takeoff. Likewise, the early ankle moment differences reversed 500 to 400 ms before takeoff, confirmed by significantly larger late countermovement angular impulse in the FJ condition. During propulsion, the three extensor joint moments peaked roughly in a proximal-to-distal sequence (Figure 6 and Table 4). The EMG differences with jump direction for HA and RF are consistent with the peak knee extensor moment and angular impulse being higher in VJ (250 ± 60 Nm and 50.6 ± 1.1 Nm s) com-

13 Muscle Activity During Jumping 217 Figure 6 Ensemble time histories of hip, knee, and ankle moments (Nm). Backward jump (thin solid line), vertical jump (thick solid line), intermediate jump (crosses), and forward jump (open circles) conditions are plotted from quiet stance to takeoff. Positive moments represent extensor moments. Horizontal bars represent average duration of countermovement (light gray) and propulsion (dark gray) phases for the 12 jumpers. pared to FJ (230 ± 46 Nm and 43.3 ± 0.8 Nm s). Peak hip moments were similar across jumping conditions, but the BJ condition displayed a prolonged extensor pattern that resulted in a larger angular impulse (Table 4). The extensor joint moments fell sharply as propulsion continued, consistent with the fall in muscle activity. The hip moment in BJ remained extensor, but in the other jumps it either fell to zero (VJ) or became flexor (FJ and IJ) in the final 75 ms before takeoff. The knee moments also became flexor just before takeoff (Figure 6), earliest in the FJ condition that displayed a strong drop in RF and VL activity and reduced angular impulse. The peak ankle plantarflexor moment and angular impulse were lowest in the BJ condition, consistent with reduced GA activity. The ankle plantarflexor moment patterns were almost identical in the final 30 ms before takeoff, even though the peak occurred significantly earlier during FJ.

14 218 Jones and Caldwell Table 4 Mean (± SD) Values for Peak Joint Moments and Their Timing, and Angular Impulse for the 3 Jump Phases BJ VJ IJ FJ p-values Peak moment (Nm) Hip 285 ± ± ± ± Knee 235 ± 52 a,b 250 ± 60 a 234 ± 58 a,b 230 ± 46 b 0.05* Ankle 206 ± 43 b 225 ± 51 a 236 ± 52 a 233 ± 55 a 0.01* Time to peak moment (ms) relative to takeoff Hip 227 ± 66 b 275 ± 75 a 293 ± 55 a 289 ± 54 a 0.01* Knee 215 ± ± ± ± Ankle 145 ± 63 b 143 ± 75 b 152 ± 74 b 205 ± 69 a 0.01* Relative H = K > A H = K > A H > K > A H > K = A 0.01* Early countermovement phase angular impulse (Nm s) Hip 14.4 ± ± ± ± Knee 10.9 ± ± ± ± Ankle 7.1 ± 4.4 a 6.0 ± 3.7 a,b 6.7 ± 4.7 a,b 4.4 ± 4.8 b 0.02* Late countermovement phase angular impulse (Nm s) Hip 31.9 ± 9.6 b 29.3 ± 10.3 b 30.2 ± 13.0 b 36.6 ± 11.9 a 0.01* Knee 23.8 ± ± ± ± Ankle 18.2 ± 8.1 b 16.3 ± 6.0 b 17.5 ± 6.8 b 22.8 ± 8.7 a 0.01* Propulsion phase angular impulse (Nm s) Hip 55.0 ± 18.0 a 48.1 ± 19.1 b 44.1 ± 18.9 b,c 39.0 ± 15.1 c 0.01* Knee 45.2 ± 13.0 b,c 50.6 ± 12.7 a 47.8 ± 12.0 a,b 43.3 ± 10.0 c 0.01* Ankle 45.3 ± 12.0 b 54.3 ± 15.0 a 56.7 ± 13.6 a 54.4 ± 13.2 a 0.01* Note: The p-values indicate statistical significance of main effect of takeoff angle. Superscripts indicate conditions that were significantly different (i.e., a b c d at p < 0.05). Final row of timing data indicates relative time to peak moments between hip (H), knee (K), and ankle (A) joints from earliest to latest peaks (e.g., H > K refers to H peaking earlier than K). Discussion Differences in monoarticular and biarticular muscle usage have been well documented for different multijoint movements (Bobbert & van Zandwijk, 1994; de Boer, Cabri, Vaes, et al., 1987; Jacobs, Bobbert, & van Ingen Schenau, 1996; van Ingen Schenau, 1989; van Ingen Schenau et al., 1987; van Soest et al., 1993). Our aim was to study mono- and biarticular muscle activity during countermovement jumps in different directions. We hypothesized that monoarticular muscle usage would be similar in all jump directions because of the common need to generate

15 Muscle Activity During Jumping 219 energy during lower extremity extension (Jacobs & van Ingen Schenau, 1992; van Ingen Schenau, 1989; van Ingen Schenau et al., 1992; 1995). Biarticular muscle activity patterns were expected to vary due to a possible conflict between the tasks of energy transport across adjacent joints and GRF direction control. Our results should be interpreted within the usual assumptions concerning the use of EMG, planar analysis, and inverse dynamics computations (Winter, 1990). One limitation is that an EMG represents a control signal rather than a muscle force output. Muscle forces and powers could be estimated using models that account for their force-velocity, force-length, and elastic characteristics (e.g., Caldwell, 1995). However, the EMG signals can be used to assess how the central nervous system coordinates muscle activation. Caution must be employed in comparing EMG data with kinematic and kinetic variables, due to an electromechanical delay of 10 to 100 ms (Corcos, Gottlieb, Latash, Almeida, & Agarwal, 1992; Komi, 1979). The planar analysis is justified by the absence of significant mediolateral motion in these sagittal-plane jumps. Finally, while inverse dynamics produces only net joint moments, the EMG data allow some assessment of the specific muscles responsible for these moments. For example, larger changes are seen between jump conditions in our EMG data than in the joint moments, most likely because of changes in the degree of antagonistic co-contraction and in the activity of muscles not monitored (e.g., iliopsoas). In general the EMG, kinematic, and joint moment patterns in the VJ condition are consistent with other studies (Bobbert & van Ingen Schenau, 1988; de Graaf, Bobbert, Tetteroo, & van Ingen Schenau, 1987). However, SO and GA activity peaked earlier and fell prior to takeoff for the jumpers in our study, probably due to differences in skill between participant populations. These other studies used more skilled jumpers who may have delayed plantarflexion in order to maximize CM takeoff speed. The few studies on jumping in directions other than vertical include Jensen and Phillips (1991), who emphasized the timing of joint coordination, and Ridderikhoff et al. (1999), who studied squat jumps that lacked a countermovement. One salient feature in the latter study was a distinct change in lower extremity joint moments in vertical vs. forward jumps, contrasting with our much more visually similar moments. In the present study, the hypothesis of similar monoarticular muscle usage in each jump direction was supported by the lack of statistical differences in peak EMG levels, but not by the average EMG data or qualitative analysis of activity patterns. The knee extensor VL had reduced average activity in the FJ condition, with a sharp reduction in activity in the last 100 ms before takeoff (Figure 4). Also, GM and TA exhibited temporal changes with jump direction that served to address the unique requirements of specific directional jumps. Extended GM activity helped control HAT extension in the BJ, while earlier TA activity in the FJ initiated horizontal CM motion to help attain the correct direction. In contrast, our hypothesis of biarticular activity changes with takeoff angle was supported by statistical differences in EMG for HA, RF, and GA. The most distinct change was in HA activity, which increased sequentially as the jump direction changed from backward (BJ) to upright (VJ) to more forward (IJ, FJ). Its antagonist RF demonstrated reciprocal alterations with direction, although these changes were not as distinct. Conflicts between energy transfer and GRF directional control were avoided by initiating and developing the appropriate jump direction during the countermovement. The CM trajectory was initiated by early TA and RF activity that accel-

16 220 Jones and Caldwell erated the body CM downward, and in some cases forward. TA activity was earlier and stronger in FJ and IJ, therefore increasing horizontal momentum in these forward jumps. Comparison of linear forward momentum at the end of the countermovement between conditions confirms the notion that these early EMG differences were important for directional control. Strong early TA activity was contraindicated in BJ and VJ conditions, as it would bring the CM too far forward and would require subsequent posterior acceleration. In FJ, the early TA and RF activity led to reduced ankle plantarflexor and elevated hip flexor moments, resulting in increased leg and HAT rotation. Around the same time, a slightly larger knee extensor moment prevented excessive knee flexion, resulting in forward translation of the body CM as the upper body was catapulted forward during the countermovement. Because the propulsion phase began from a unique position with angular and linear momenta already initiated for the required jump direction, the joint moments during propulsion could be relatively similar in each jump. This contrasts with squat jumps in which both energy generation and directional control must be attained solely in the propulsion phase, and thus distinct joint moment patterns are needed for different jump directions (Ridderikhoff et al., 1999). The early EMG, kinematic, and kinetic differences between jumping directions suggest that the countermovement was an integral part of the directional strategy and was not used solely to effect a stretch shorten cycle to enhance the subsequent propulsion (Asmussen & Bonde-Petersen, 1974). Furthermore, this strategy is consistent with the exploitation of movement dynamics in the production of skilled motor tasks, as suggested by Bernstein (1967). Although the relatively similar joint moment patterns indicate that basic jumping coordination was preserved, the HA and RF activity differences beginning 400 ms before takeoff are consistent with changes in thigh and leg kinematics, and with the purported directional roles suggested by Jacobs and van Ingen Schenau (1992). This late countermovement / early propulsion phase appears to be a critical transition between the establishment of correct jump direction and the generation of CM speed, and is concurrent with the largest differences in knee and ankle moments. However, EMG differences may also be associated with specific tasks that differ with jump condition, such as controlling HAT rotation. The sequential increase in HA activity with more forward jumps may be due to the need to resist the larger gravitational torques caused by the more forward-angled HAT. The tendency for peak GM activity to occur earlier in FJ and IJ support this possibility, as does the reduction in hip flexor RF activity in FJ. Another direction-specific task is controlling the position of the GRF center of pressure (COP). In the forward jump (and to a lesser extent the IJ and VJ), ankle plantarflexion was effective in generating forward CM motion because the CM had been positioned anterior to the fifth metatarsal head by the earlier catapulting action (Figure 3). In contrast, for BJ the CM was always posterior to the fifth metatarsal, and the foot remained stationary until 50 ms before takeoff when it left the ground due to the body s momentum. Strong plantarflexion before that time would have moved the COP forward along the base of the foot (Bobbert, Houdijk, de Koning, & de Groot, 2002). This would cause the GRF line of action to fall in front of the CM and thus promote unwanted backward angular momentum (see landing considerations below). Therefore, during BJ the reductions in GA activity

17 Muscle Activity During Jumping 221 and plantarflexor moment beginning 400 ms before takeoff help to control both the COP location and the GRF direction. In directional jumps, preparation for the subsequent landing is another potential influence. In VJ the extended and vertically oriented takeoff posture is ideal for landing safely. In FJ the forward-angled takeoff posture must be modified substantially during the flight, as the feet must move in front of the CM before touchdown. This FJ landing requirement may explain the hip and knee flexor moments and the reductions in GM, RF, and VL activity in late pushoff. Likewise, in BJ the extended GM activity and hip extensor moment in late pushoff helps backward thigh rotation to move the feet in a posterior direction after takeoff. Thus, for safe landing in backward jumps, one must avoid excessive backward total body angular momentum, which helps explain the reduced GA activity and plantarflexor torque prior to takeoff in BJ. In summary, our data suggest a strategy of muscular coordination that does not fall strictly under the monoarticular / biarticular dichotomy which we hypothesized. Activity of TA and RF and adjustment of ankle and knee moments initiate directional control early in the countermovement. The appropriate CM trajectory is implemented by coordinated muscle activity during the countermovement such that directional control of the jump has begun before propulsion begins. GM, HA, and RF act to control the massive HAT segment through the critical transition from countermovement to propulsion, while CM velocity is increased through the sequential action of the monoarticular VL and SO during propulsion. Late in propulsion, the GM, RF, VL, GA, and TA activation levels are tuned to adjust the degree of forward or backward angular momentum and foot position in preparation for landing. This muscular coordination strategy enables maximal energy generation and transfer between joints for propulsion without interference from the task of controlling the direction of the jump. References Asmussen, E., & Bonde-Petersen, F. (1974). Storage of elastic energy in skeletal muscles in man. Acta Physiologica Scandinavica, 91, Bernstein, N.A. (1967). The coordination and regulation of movements. London: Pergamon. Bobbert, M.F., Houdijk, H., de Koning, J.J., & de Groot, G. (2002). From a one-legged vertical jump to the speed-skating push-off: A simulation study. Journal of Applied Biomechanics, 18, Bobbert, M.F., & van Ingen Schenau, G.J. (1988). Coordination in vertical jumping. Journal of Biomechanics, 21, Bobbert, M.F., & van Zandwijk, J.P. (1994). Dependence of human maximum jump height on moment arms of the bi-articular m. Gastrocnemius; a simulation study. Human Movement Science, 13, Bresler, B., & Frankel, J.P. (1950). The forces and moments in the leg during level walking. Transcripts of the American Society of Mechanical Engineers, 72, Caldwell, G.E. (1995). Tendon elasticity and relative length: Effects on the Hill two-component muscle model. Journal of Applied Biomechanics, 11: Corcos, D.M., Gottlieb, G.L., Latash, M.L., Almeida, G.L., & Agarwal, G.C. (1992). Electromechanical delay: An experimental artifact. Journal of Electromyography and Kinesiology, 2,

18 222 Jones and Caldwell De Boer, R.W., Cabri, J., Vaes, W., Clarijs, J.P., Hollander, A.P., de Groot, G., & van Ingen Schenau, G.J. (1987). Moments of force, power and muscle coordination in speed skating. International Journal of Sports Medicine, 8, De Graaf, J.B., Bobbert, M.F., Tetteroo, W.E., & van Ingen Schenau, G.J. (1987). Mechanical output about the ankle in countermovement jumps and jumps with extended knee. Human Movement Science, 6, Dempster, W. (1955). Space requirements of the seated operator (USAF, WADC, Tech. Report ). Wright Patterson Air Force Base, OH. Jacobs, R., Bobbert, M.F., & van Ingen Schenau, G.J. (1996). Mechanical output from individual muscles during explosive leg extensions: The role of biarticular muscles. Journal of Biomechanics, 19, Jacobs, R., & van Ingen Schenau, G.J. (1992). Control of an external force in leg extensions in humans. Journal of Physiology, 457, Jensen, J.L., & Phillips, S.J. (1991). Variations of the vertical jump: Individual adaptations to changing task demands. Journal of Motor Behavior, 23, Komi, P.V. (1979). Neuromuscular performance: Factors influencing force and speed production. Scandinavian Journal of Sports Science, 1, Ridderikhoff, A., Batelaan, J.H., & Bobbert, M.F. (1999). Jumping for distance: Control of the external force in squat jumps. Medicine and Science in Sports and Exercise, 31, Selbie, W.S., & Caldwell, G.E. (1998, August). Common features of simulated jumping. In Proceedings of the North American Congress on Biomechanics III (pp ). Waterloo: University of Waterloo Van Ingen Schenau, G.J. (1989). From rotation to translation: Constraints on multi-joint movements and the unique action of bi-articular muscles. Human Movement Science, 8, Van Ingen Schenau, G.J., Bobbert, M.F., & Rozendal, R.H. (1987). The unique action of biarticular muscles in complex movements. Journal of Anatomy, 155, 1-5. Van Ingen Schenau, G.J., Boots, P.J.M., de Groot, G., Snackers, R.J., & Van Woensel, W.W.L.M. (1992). The constrained control of force and position in multi-joint movements. Neuroscience, 46, Van Ingen Schenau, G.J., Dorssers, W.M.M., Welter, T.G., Beelen, A., de Groot, G., & Jacobs, R. (1995). The control of mono-articular muscles in multijoint leg extensions in man. Journal of Physiology, 484, Van Soest, A.J., Schwab, A.L., Bobbert, M.F., & van Ingen Schenau, G.J. (1993). The influence of the biarticularity of the gastrocnemius muscle on vertical-jumping achievement. Journal of Biomechanics, 26, 1-8. Winter, D.A. (1990). Biomechanics and motor control of human movement. New York: Wiley.