Foot placement modifies kinematics and kinetics during drop jumping

Abstract

Foot placement modifies kinematics and kinetics during drop jumping. Med. Sci. Sports Exerc., Vol. 31, No. 5, pp. 708-716, 1999:

Purpose: 

Sprinting, bouncing, and spontaneous landings are associated with a forefoot contact whereas walking, running, and jumping are associated with heel-toe foot placement. Because such foot placement strategies influence landing mechanics or the ensuing performance, the purpose of this work was to compare lower extremity kinematics and kinetics and muscle activation patterns between drop vertical jumps performed with heel-toe (HTL) and forefoot (FFL) landings.

Methods: 

Ten healthy male university students performed two types of drop jump from a 0.4-m high box placed 1.0-m from the center of the force plate. They were instructed to either land first on the ball of the feet without the heels touching the ground during the subsequent vertical jump, i.e., forefoot landing jump (FFL), or to land on the heels followed by depression of the metatarsals, i.e., heel-toe landing jump (HTL). Three successfully performed trials per jump type were included in the analysis. The criteria for selection of the correct jumps was proper foot position at contact as judged from video records and the shape of force-time curve.

Results: 

The first peak and second peak determined from the vertical force-time curves were 3.4 times greater and 1.4 times lower for HTL compared with those with FFL (P < 0.05). In the flexion phase of HTL, the hip and knee joints contributed 40% and 45% to the total torque, whereas during FFL the greatest torque contributions were 37% for both the knee and ankle joints. During the extension phase, the greatest torque contributions to the total torque were 41% and 45% by the knee and ankle joints during HTL and 34% and 55% during FFL. During the flexion phase, power production was 20% greater (P < 0.05) in HTL than in FFL, whereas during the extension phase power production was 40% greater in FFL than in HTL. In the flexion phase of HTL the hip and knee joints produced the greatest power, and during the extension phase the knee and ankle joints produced the greatest power. In contrast, during both the flexion and extension phases of FFL, the knee and ankle joints produced the greatest power. The EMG activity of gluteus, vastus lateralis, and plantar flexor muscles was similar between HTL and FFL in most cases except for the greater vastus lateralis EMG activity during precontact phase in HTL than in FFL and the greater gastrocnemius activity in FFL than in HTL.

Conclusion: 

Foot placement strategy modifies the individual joint contributions to the total power during drop jumping.

Landing after vertical or vertical-horizontal fall and the collision of the human body with the ground is a frequent occurrence in human locomotion. Indeed, since force platforms were first used to study ground reaction forces ⁽¹⁸⁾, landing has been studied extensively in biomechanics ^{(3-6,15-18,21,22,26,33,35,36)}. Nonetheless, landing presents a paradox concerning the selection of foot placement strategy during locomotion. Sprinting, bouncing, and spontaneous landings are associated with forefoot contact ^{(10,24,25,27,30)} whereas walking, running, and jumping in track and field are frequently associated with heel-toe foot placement ^{(12,14,19,28-30)}. One functional benefit of the forefoot strategy is shock attenuation and energy absorption that protects bones and joints superior to the site of impact from excessive forces by loading the muscles of ankle joint in particular ^(1,4,6,21). A second benefit is an enhanced muscle shortening caused be the elastic energy release from the previously stretched plantar flexors. In contrast, the heel-toe placement strategy is associated with less shock absorption and minimal eccentric function of the plantar flexors. The heel-toe strategy may be superior, however, in partitioning the horizontal momentum into larger vertical and smaller horizontal momenta, i.e., ankle joint plays a significant role in energy absorption and generation and influences the whole ground reaction force-time history ⁽¹⁾ and may therefore critically affect jumping performance.

Previous studies have examined the effects of landing surfaces ⁽²²⁾, the manipulation of stiffness ^{(15,17,21,26,33,35,36)}, or elastic energy storage and utilization and coordination ^(2-6,17) on lower extremity joint kinetics, but because these activities were not followed by jumping, controlling foot placement strategy was not an issue. In most studies on landing or drop jumping, subjects landed on the metatarsals or used a metatarsal followed by heel contact strategy. Soft and stiff forefoot landings have been identified on the basis of joint angle position at the instant of touchdown ⁽¹⁶⁾ or the range of angular displacement during landing ^{(15,20,33,35)}. Whereas in stiff landings the summed joint work was dominated by the ankle joint, in soft landings the work was distributed more uniformly across the joints ⁽¹⁵⁾. Fukuda ⁽²⁰⁾ found hip and knee dominance in shock attenuation. Bobbert et al. ^(3,4) reported ground reaction force data and lower extremity joint moments and powers by applying countermovement and bounce type drop jumps falling from several heights. They observed ankle joint dominance in power production for bounce drop jump in contrast with countermovement drop jump where knee joint dominance was observed ⁽⁴⁾. These studies showed that all lower extremity joints participated in energy absorption and generation in landings and jumps using forefoot landing strategy, and as landing height increased the role of the ankle musculature increased. Interestingly, while many athletes perform drop jumping and plyometric drills with both types of foot placement strategies as a training modality in various sports (track, ball games), the two strategies have not been compared critically.

Therefore, the purpose of the present study was to compare lower extremity kinematics, kinetics, and muscle activation patterns between heel-toe and forefoot landing strategies. Such information is of interest not only for athletes to improve training exercises but also for clinicians because certain gait abnormalities are associated with toe-heel and not the expected heel-toe foot strike as in idiopathic toewalking in cerebral palsy ⁽³¹⁾. The drop jump was used as a model because it provides a well controlled experimental approach to investigate the difference in energy absorption and eccentric-concentric muscle function between forefoot and heel-toe landings.

METHODS

Subjects. Ten healthy male university students (age: 23.5 ± 2.5 yr; body mass: 82.5 ± 4.6 kg; and height: 1.85 ± 0.06 m) volunteered as subjects for this study. Before the testing subjects signed an informed consent form approved by the Policy and Review Committee on Human Research of East Carolina University.

Instrumentation. A 1.2 m by 0.6 m AMTI force platform (LG-6-4-1, AMTI, Newton, MA) located in center of a walkway was used to measure ground reaction forces at 1 kHz. A NAC high speed video camera (V-14, NAC Visual system, Woodland Hills, CA) was positioned 10 m from the force platform to videotape the drop jumps in the sagittal plane at 0.2 kHz. The field of view was 2 m by 2.4 m. Video records were digitized with a Peak Performance system (Englewood, CO). Surface EMG activities of the gluteus maximus, vastus lateralis, and lateral head of gastrocnemius were recorded with the Noraxon telemetric system (Scottsdale, AZ). The three EMG and three force platform signals (F_z, F_x, M_y) were collected with the Myosoft software package at 1 kHz (Noraxon, Scottsdale, AZ).

Protocol. Body mass and girth of upper thigh, knee, ankle, and metatarsal head for each limb were measured. Reflective markers were placed on the lateral side of the shoe at fifth metatarsal head and heel, and on the surface of the lateral malleolus, femoral condyle, greater trochanter, and shoulder on the side of the body that faced the camera. A marker placed on the front corner of the force platform was used to convert the center of pressure data from a force platform based system to the laboratory reference frame (see below).

In one session, each subject performed two types of drop jump from a 0.4 m high box placed 1.0 m from the center of the force plate. Subjects were instructed to land either first on the ball of the feet without the heels touching the ground during the subsequent vertical jump, i.e., forefoot landing jump (FFL) or to land on the heels followed by depression of the metatarsals, i.e., heel-toe landing jump (HTL). Trials attempted as forefoot landing jumps were not used for analysis if heels contacted the ground. Subjects inexperienced with a specific landing strategy performed several practice trials from 0.2 m and 0.4 m heights. Subjects placed their hands on the hips, pushed off from the box with one leg, closed the legs in the air, and landed on the force platform with both feet simultaneously. One leg push-off was applied to eliminate the vertical trajectory of center of mass during drop from the box and to provide a more stable vertical trunk position since the box was placed 1.0 m from the center of the force platform. Subjects were instructed to avoid large joint excursions during landing and to rebound from the platform as fast as possible. Three successfully performed trials per jump type were included in the analysis. The criteria for selection of the correct jumps was: proper foot position at contact as judged from video records and the shape of force-time curve.

Data analysis. Peak vertical forces and the corresponding times were determined from ground reaction force-time curves. Vertical impulse and mean anterior horizontal forces were calculated (Fig. 1.) Rate of force development (RFD) was calculated by dividing peak force (F_v1 and F_v2) by the time elapsed from the onset of increasing force to the peak.

Figure 1: Representative vertical (F_v) and horizontal (F_h) ground reaction force-time curves for heel-toe landing (HTL) and forefoot landing (FFL). Negative and positive values represent anterior and posterior horizontal forces. Rectified and smoothed EMG curves indicate electrical activity of the gluteus maximus (Glut), vastus lateralis (VL), and gastrocnemius lateralis (Gas) muscles during 80 ms before touchdown and during contact phase. Negative and positive times indicate descent and floor contact phases.

Reflective markers were digitized 80 ms before contact until 80 ms after take-off from the force platform to improve the accuracy of the data near the performance boundaries ⁽¹⁵⁾. Kinematic data were smoothed with an interactive cubic spline. Three kinematic dependent variables were computed at the ankle, knee, and hip joints: joint position at ground contact, and angular displacements during the flexion and extension phase (Fig. 2). Duration of joint flexion and extension was determined for each joint.

Figure 2: Joint position (θ), joint moment of force (M), and power (P) curves from representative heel-toe (HTL) and forefoot (FFL) foot strike patterns. Solid lines represent HTL and dashed lines FFL. Negative and positive values on power curves indicate energy absorption (joint flexion) and energy generation (joint extension). Negative and positive times indicate descent and floor contact phases.

Segmental masses, the mass center location of the lower extremity, and their moment of inertia were estimated using four segmental mathematical model ^(15,25), segmental masses reported by Dempster ⁽¹³⁾, and the individual subjects' anthropometric data. Center of pressure was calculated from ground reaction forces, and it was expressed initially as a distance from the center of the platform. These data were then converted into the laboratory reference frame based on the digitized location of the force platform. Often, a few center-of-pressure values at the start and end of the movement have large errors because of the low amount of ground reaction force at these times The values in the initial and final phases in the center of pressure data were compared with the location of the foot, and any center of pressure values falling behind the heel or in front of the toe were adjusted to the heel and toe positions. Approximately six center of pressure values required adjustment in each trial. Joint torques were calculated through an inverse dynamic analysis combining the anthropometric, kinematic, and kinetic data. Joint power was computed as the product of joint torque and angular velocity. Kinetic dependent variables were: joint moment, power output (Fig. 2), and work done during flexion and extension phases at ankle, knee, and hip joints

EMG data analysis included the inspection of the data for movement artifact and baseline shift. Signal were then rootmean-square processed (RMS) using 20-ms moving average. Peak RMS EMG values (in mV) were manually digitized during the descent (80 ms before landing) and during flexion and extension phases of the jump.

Statistical analyses. The dependent variables were compared with a paired two-tailed t-test between two conditions of jumping. The level of significance was set at P < 0.05.

RESULTS

Ground reaction forces. Two peaks were observed in the vertical force-time curves and labeled the first (F_v1) and second (F_v2) peak forces (Fig. 1). The vertical force increased sharply from the instant of foot contact and reached F_v1 in 0.015 s and 0.017 s in HTL and FFL, respectively. F_v1 was 3.4 times greater in HTL (P < 0.05) than in FFL (Table 1) and therefore the rate of force development was 3.8 times greater (P < 0.05) in HTL (HTL: 6613.3 ± 826.6 N·kg⁻¹·s⁻¹, FFL: 1723.5 ± 155.2 N·kg⁻¹·s⁻¹). The decrease in force after F_v1 was more pronounced in HTL. F_v2 was 2.4 times lower than F_v1 in HTL whereas F_v2 was 2.1 times greater (P < 0.05) than F_v1 in FFL. F_v2 was 1.4 times greater (P < 0.05) in FFL than in HTL.

TABLE 1: Mean and SD (in parenthesis) of variables determined from ground reaction force-time curves.

The total contact time (time elapsed from touchdown to the end of takeoff) was 1.2 times longer and the time between F_v1 to F_v2 was 1.5 times longer (both P < 0.05) in HTL compared with that in FFL (Table 1).

Joint kinematics. Subjects contacted the floor with different angular positions at the ankle and hip joints in the two conditions (P < 0.05). The ankle joint was dorsiflexed 30.1 ± 3.8° and plantarflexed 10.8 ± 2.1 at the instant of floor contact in HTL and FFL, respectively (Table 2, Fig. 2). The hip joint was flexed 5.4 ± 0.7° more in HTL compared with that in FFL. The angular position at the knee was statistically identical between conditions.

TABLE 2: Mean and SD (in parenthesis) of variables from joint kinematic analysis.

The angular displacements were also significantly different between conditions in both flexion and extension phases. There was 8.2 times more ankle dorsiflexion during FFL than during HTL. There was 1.2 and 2.0 times more knee and hip flexion during HTL than FFL (all P < 0.05). There also was 1.1 and 1.3 times more extension at the knee and hip, respectively, during the extension phase (P < 0.05). The amount of ankle plantarflexion was nearly identical between conditions during the extension phase.

Joint action synchronization. Time from first contact to maximum joint flexion was consistently the shortest for hip joint and the longest for knee joints in both HTL and FFL (Table 2). Ankle dorsiflexion and knee flexion was shorter in FFL than in HTL (P < 0.05). The extension time was the longest for hip joint. The time of extension was shorter in FFL than in HTL at all joints (P < 0.05). When the shortest time of flexion phase for a joint, that is, always the hip joint, is subtracted from the longest flexion time for a joint (knee joint in our study), a time interval is calculated that indicates a period of time when the transition from flexion to extension takes place at each joint. Table 2 shows that the calculated time interval is 2.0 times longer (P < 0.05) for HTL (0.032 ± 0.003 s) compared with that for FFL (0.016 ± 0.003 s).

Joint kinetics. The sum of the joint torques was similar between HTL (389.4 ± 89.5 N·m) and FFL (393.0 ± 112.3 N·m) during flexion phase. In the extension phase, the sum of the joint torques was greater in FFL (393.2 ± 64.2 N·m) than in HTL (317.9 ± 63.7 N·m). There was 1.3 and 1.1 times greater mean torque at the hip and knee in HTL than in FFL, and the mean torque about the ankle was 1.7 times greater in FFL than in HTL over the entire landing and jump (P < 0.05). The mean torque at the hip and knee were not statistically different between HTL and FFL during the extension phase (Table 3). The mean torque at the ankle was 2.1 and 1.5 times greater in FFL compared with that in HTL (P < 0.05) during the flexion and extension phases.

TABLE 3: Mean and SD (in parenthesis) values of torque, power, and work production for hip, knee, and ankle joint in heel-toe (HTL) and forefoot (FFL) landing.

The relative contribution of the individual joint torques to the total torque produced in the extremity was assessed by the ratios of the mean joint torques to the sum of the three mean joint torques. In the flexion phase of HTL, the largest torques came from the hip (37.3 ± 4.6% of total) and knee (44.8 ± 3.2%) joints and from the knee and ankle joints (37.3 ± 2.8 and 37.3 ± 3.3%) in FFL. In the extension phase, the largest torques came from the knee and ankle joints both in HTL (41.4 ± 3.8 and 45.4 ± 4.5%) and FFL (33.5 ± 3.5 and 55.0 ± 4.8%).

Joint energetics. Most of the joint power and work variables were significantly different between jumping techniques (Table 3). The pattern of mean power and work differences during the flexion phase was similar to the pattern of torque differences, emphasizing hip and knee roles in HTL and ankle roles in FFL. The between-subject variability of hip power data was high, indicating the unstable function of the last link in the kinematic chain. The summed muscular powers were −2605.6 ± 565.7 and −2191.7 ± 389.1 W in HTL and FFL during the flexion phase. The total work during the flexion phase was −167.1 ± 48.4 and −142.9 ± 32.2 J in HTL and FFL. Hip and knee had greater negative power and work in HTL compared with that in FFL, whereas the ankle joint had greater value at FFL during this phase. In the extension phase, the sum of joint power was 2451.8 ± 464.7 and 3515.2 ± 368.9 W in HTL and FFL. Total work during the extension phase was 327.4 ± 72.3 and 454.7 ± 56.7 J in HTL and FFL. Greater work was calculated about knee and ankle joints during the extension phase in FFL than in HTL.

EMG activity. During the precontact phase vastus lateralis displayed higher EMG activity in HTL compared with that in FFL (P < 0.05). In contrast, EMG activity was elevated (P < 0.05) for gastrocnemius lateralis in FFL. During the flexion phase, the EMG activity of the three muscles was similar in both HTL and FFL. However, the EMG activity was enhanced compared with the precontact phase for all muscle groups except for the gastrocnemius lateralis in FFL. For the extension phase there were no significant differences in EMG activity in the three muscles between HTL and FFL, except for the vastus lateralis in FFL (Table 4, Fig. 1). Hip extensor EMG activity remained almost unchanged compared with that of flexion phase. Reduced vastus lateralis EMG activity was observed in both HTL and FFL during extension phase whereas gastrocnemius muscle EMG activity was greater during the extension phase of FFL compared with that during HTL.

TABLE 4: Mean and SD (in parenthesis) of average EMG amplitudes for muscles studied. The values are expressed in terms of the average levels attained during eccentric phase for HTL.

DISCUSSION

Comparison of HTL and FFL force-time curves. There were distinct differences in the ground reaction force-time profiles between HTL and FFL. Heel-toe landing resulted in large and rapid (20-25 ms) vertical impact forces (F_v1) associated with the stiffer landing mediated by the skeletal system of the lower extremities. In HTL, the ankle plantar flexors could not attenuate the impact because the center of pressure is located at the calcaneus and closer to ankle joint center than in FFL. Accordingly, F_v1 and RTD were 3.4 and 3.8 times greater in HTL than in FFL-an observation also reported previously ⁽³⁵⁾. A second difference between HTL and FFL was related to the second peak in the vertical ground reaction force, Fv₂. After F_v1, vertical force declined sharply because of the flexion of the hip and knee joints followed by a second increase in the vertical force, F_v2, associated with the decreasing rate of flexion of the hip, knee, and ankle joints. This phenomenon was observed in many landing and drop jump studies ^{(3-6,11,15-17,21,26,35,36)}. F_v2 was 2.4 times lower and 2.1 times higher than F_v1 in HTL compared with that in FFL, indicating that in HTL part of the kinetic energy was applied to the passive motion system (bones, joints, ligaments, cartilage).

The force-time history of HTL observed in the present study was similar to the force-time profile reported previously during high jump, long jump, and triple jump tests in which ground contact occurs with a heel-toe pattern ^{(11,19,29,30,32)}. The force-time curves of FFL in the present work resembled the force-time curves reported in studies that examined the mechanism of shock attenuation ^{(11,15-17,21,22)} or the efficiency of drop jumps ^(3-5,8,9) using forefoot landing. Specifically, Bobbert et al. ⁽⁴⁾ compared countermovement and bounce type drop jump. Their maximum vertical forces (F_v2) in the bounce drop jump were considerably less than ours. The differences perhaps result from the higher dropping height, the twice shorter time to F_v2, and the smaller hip-knee flexion and dorsiflexion during the flexion phase in the present study. Two peaks in vertical ground reaction force were also determined in studies that compared soft and stiff landings ⁽¹⁵⁾ with toe-heel and toe (metatarsal) strategies ^(17,21,22). When the heels were allowed to contact the ground after forefoot touchdown, the first peak was always significantly lower compared with the second peak as was observed in FFL. However, it should be noted that the second peak in shock attenuation studies ^{(11,15-17,21,26,35,36)} can be considered as a passive mechanism response ⁽²⁸⁾ as a consequence of hard heel contact which is not the case in the present study.

The active peak force (F_v2) was 1.6 times greater (P < 0.05) in FFL than in HTL because of the more synchronized action of the three joints in this condition. The flexion phase in all three joints was completed in 0.016 s in FFL and in 0.032 s in HTL. This more simultaneous ending to the flexion phase at each joint in FFL produced a more rigid body and therefore a higher vertical peak force which occurred nearly at the transition between flexion and extension phases. The shorter transition time between flexion and extension phases in FFL compared with that in HTL indicated that FFL was a more explosive jumping movement compared with the more drawn out motion in HTL. The joint motions were more synchronized (extension started for each joint within a shorter time of period) in the present study, and particularly in FFL, than in Bobbert et al. ⁽⁴⁾ which may account for the greater vertical reaction forces in this study.

Comparing HTL and FFL force-time history with a mechanical model. Examination of the ground reaction force-time histories of HTL and FFL leads to a conclusion similar to ones reached by Alexander et al. ⁽¹⁾. By applying a mass-spring model there was only one peak in the midpart of the curve that was similar to the FFL curves except the first, low peak in our case. By adding a smaller mass to the distal end of the spring that directly contacts the surface after free fall, the resultant force-time curve has a high first peak followed by a lower second force peak with a lower loading rate than in the case of the first peak. This force-time curve resembles the HTL curves obtained in our study. In another model of Alexander et al. ⁽¹⁾ the small mass was connected to spring-dashpot combination that directly contacted the ground. The resultant force-time curve showed an initial impact force peak with some subsequent damped vibration, followed by an active force peak. This type of force-time curve is the most similar to our FFL curves except the damped vibration. On the basis of these similarities one may assume that the trunk represents the first mass and the hip extensors act as the initial spring, the thigh is the second mass and knee extensors are the second spring. Developing this train of ideas, a third mass and spring or spring dashpot component, which are the lower leg and plantar flexors, should be added to the model. A three-component model may reduce the vibration content of the force-time curve. This analogy supports our explanation concerning the reasons of the difference in vertical forcetime curves between HTL and FFL and points out the significance of the functional order and the role of the joint action and body segments as it was demonstrated in a mathematical model ⁽³⁴⁾.

Joint kinematics. Soft and stiff landings ^{(15,33,35,36)} and counter movement and bounce type drop jumps ^(3,4) have been differentiated on the basis of angular displacement during the flexion phase: the larger the joint flexion is the softer the landings. Both HTL and FFL can be considered partly stiff and partly soft concerning ankle joint action because of heel touchdown and small angular displacement in HTL and large angular displacement in FFL during flexion phase. On the contrary, in HTL hip and knee angular displacements are significantly larger compared with those in FFL. Interestingly, the summed angular displacement of the three joints is significantly larger for FFL than for HTL because of the 8.2 times larger ankle joint angular displacement. However, both HTL and FFL are bounce-type rather than counter-movement type drop jumps since the minimum joint angle at the transition from flexion to extension phase for all joints is greater than that reported by Bobbert et al. ^(3-5) for bounce-type drop jumps. The duration is significantly longer for all joints studied in HTL compared with that in FFL, but the times for both HTL and FFL are much shorter than those of Bobbert et al. ⁽³⁾.

Joint kinetics and energetics. A critical difference between the two techniques was the minimal energy absorption by the plantarflexors in HTL. On average the negative work done at the ankle was only −15.7 ± 5.4 J in HTL which was 23% of the corresponding work in FFL. The reduced ankle work contributed to the increased rigidity and greater net vertical impulse in HTL during the flexion phase. As a consequence, the significance of the knee and hip extensors increased in HTL as indicated by the greater torque, power, and work values at these joints during the flexion phase. Bobbert et al. ⁽⁴⁾ reported that the highest mean power was produced by the ankle plantarflexors followed by knee and hip extensors while using a forefoot contact strategy; these results are similar to ours for the flexion phase in FFL. In contrast, HTL had the highest mean power produced by the knee extensors followed by the hip extensors and ankle plantarflexors. The present mean power values were similar to those reported previously ⁽⁴⁾. DeVita and Skelly ⁽¹⁵⁾ reported the highest mean torque and power values at the hip joint in soft and stiff landings from 0.60 m height while Zhang et al. ⁽³⁶⁾ and Bobbert et al. ⁽⁵⁾ reported that the role of the knee joint increased with increased falling heights. The results of this study are in a good agreement with the previous studies indicating that the significance of each joint's contribution to the total power increases or decreases as a function of the landing techniques and the force applied to the body.

The present results demonstrate the importance of separating the flexion and extension phases in jumping research. The relative and absolute contributions of torque, power, and work were different at each joint in flexion and extension phases. The total work at the hip, knee, and ankle joints was 16.9 ± 4.3% greater during the flexion phase in HTL compared with that in FFL. In contrast, the total work done was 38.9 ± 4.1% greater in FFL compared with that in HTL during the extension phase. The relative contributions of each joint to the total work were also different between flexion and extension phases. The hip and knee extensors contributed 26.5 ± 7.6% and 64.1 ± 5.4%, and the remaining 9.4 ± 3.2% came from the ankle plantarflexors during the flexion phase in HTL. The corresponding values in FFL were 3.2 ± 1.1, 49.2 ± 3.9, and 47.6 ± 3.8%, indicating the increased significance of the plantarflexors and the reduced importance of the hip extensors. The relative contribution of the joints to the total work during extension phase shows similarities between the conditions. In HTL condition the hip, knee, and ankle joints contributed 19.5 ± 6.1, 44.5 ± 3.5, and 36.0 ± 5.9%, respectively. The corresponding values in FFL were 14.8 ± 4.2, 48.8 ± 3.7, and 36.4 ± 4.8%. Because both in the flexion and extension phases the mean torque and power about the joints showed similar relative contributions, it is apparent that the ankle is the primary source of energy absorption during forefoot landings ⁽⁴⁾. Energy absorption was shifted to the hip joint in the heel contact technique.

Myoelectrical potentiation and elastic energy. During the negative phase of jumping, kinetic energy can be used for enhancing muscle force which becomes visible in the increasing electrical activity of the muscles ⁽⁹⁾. A part of the energy is believed to be stored as elastic energy in the series elastic elements of the muscles ^(8,23). During the extension phase of vertical jumping the absorbed energy can be utilized if muscles shorten immediately after negative work phase ^(8,23). If there is a long time delay between the flexion and extension phases, most of the absorbed energy dissipates as heat that has usually occurred in landing studies ^{(3-6,11,15-17,21,26,35,36)}. The elastic energy storage and utilization in vertical jump, however, has been questioned. Bobbert et al. ⁽⁷⁾ discussed several possibilities that might explain the higher work production during the extension phase if the muscle was stretched previously by using countermovement and squatting jump model. They ruled out that the enhancement of joint work was the consequence of elastic energy storage during the flexion phase because of the low stretching speed and the long transitional period from flexion to extension. However, it should be emphasized that the stretching conditions of muscles in drop jump are different from those in counter movement jump. In our study the stretching velocity was much higher than in a countermovement jump and the contact time was considerably shorter. Therefore, it can be assumed that one part of the kinetic energy was stored as elastic energy during the flexion phase and was recalled in the extension phase. We predicted that the possibility of energy storage was much less in HTL than in FFL because of the difference in foot strike pattern. In the present study an enhanced myoelectrical potentiation occurred in all muscle groups as indicated by the increased electrical activity of the muscles during the flexion phase relative to the precontact phase. In this respect there is no difference between HTL and FFL conditions. However, there is a significant difference between HTL and FFL in energy absorption since the summed power and work production were considerably greater in FFL than in HTL. The ratio between the flexion and extension phase in power and work is also higher in FFL, indicating that more work was done during the extension phase than in the flexion phase in FFL than in HTL. The larger amount of work can be attributed to the better synchronization of muscle work and to the more effective utilization of the plantar flexors. Bobbert et al. ⁽⁵⁾ reported knee extensor and plantar flexor dominance for power and work production in both the flexion and extension phases, but the relative contribution of maximum power produced by the plantar flexors has been reduced during the extension phase, which may be the consequence of the longer extension phase and the reduced torque.

We observed an increased mean power and work production for plantarflexors in FFL and HTL during the extension phase compared with that during the flexion phase, which is in a good agreement with the greater EMG activity of gastrocnemius muscles ⁽²⁾. However, the relative contribution of plantar flexors is considerably higher in FFL than in HTL. This difference cannot be related to the EMG activity since there is no difference between the two landing conditions. The question arises as to why the plantar flexors were able to produce higher joint torque, power, and work output in FFL without increasing electrical activity of gastrocnemius muscle. The most likely explanation is that elastic energy was absorbed in the elastic elements of gastrocnemius muscles and was used during the extension phase. When a previously stretched muscle shortens it is not necessary to increase the force generation of the contractile elements because the shortening elastic elements produce work without increasing EMG activity ⁽⁹⁾. We may argue that gastrocnemius muscles being bi-articular and cannot be stretched because of the knee flexion. This is very likely in HTL because angular displacement of the ankle is small compared with that in knee flexion. However, the angular displacement of the ankle joint is greater than that of the knee flexion, which may allow the elastic elements to be lengthened.

In conclusion, landing with an FFL compared with an HTL strategy during a drop jump resulted in greater energy absorption in the lower extremity musculature predominantly because of the function of the plantar flexor muscles. In addition to the re-use of this greater strain energy in FFL during the extension phase of the jump, a better synchronization of power production by the ankle, knee, and hip joints could also be responsible for the greater power production in FFL compared with that in HTL. The pattern of power production was greater distally in FFL, but it was greater proximally in HTL. Foot placement while landing during a drop jump thus clearly modifies the magnitude and distribution of power production.