Lower-limb amputees participate in the long-jump event at the highest level in international competition. Advances in prosthetic technology and improvements in training now mean that it is becoming increasingly competitive at elite level, and thus any small improvements in technique could mean the difference between winning and losing. However, what is known about the long-jump technique currently used by amputees athletes is limited, nor is it known whether the current technique is the most effective. Scant literature exists of analyses of amputee long jump (1,9,10,12); in particular, there are no published studies concerning the touch-down/take-off technique used by female amputee athletes. Lower-limb amputee athletes cannot use the same technique as able-bodied athletes because of the absence of active lower-extremity musculature and also because prosthetic limbs have different mechanical properties than intact limbs. This is further complicated for those athletes with transfemoral amputations who lack active knee-joint extensors to generate propulsive force. Furthermore, absence of neuromuscular control of prosthetic joints may make targeting the take-off board more difficult for transfemoral amputees (12). Thus, further investigation of the technique used by these athletes is required.
Much is known about able-bodied long-jump technique, for which several elements seem to be crucial in achieving optimal take-off conditions (5). The established long-jump model indicates that a relationship exists between approach speed and distance jumped (4,7,8); an increase in approach speed will result in an increase in distance jumped as long as an optimal take-off technique is used. An optimal take-off technique for able-bodied athletes has been shown to include lowering of the center of mass during the last few steps, obtaining the correct body posture at touch-down, and successfully "pivoting" over the take-off leg to generate sufficient vertical velocity while minimizing losses in horizontal velocity. The success of the pivot is partly dependent on approach speed and eccentric leg strength. Male amputee athletes have previously been reported to conform to this model (9,11), with some adjustments to the technique. However, it is not known whether female amputee athletes also conform to this model or whether they make similar adjustments in technique to male amputee long jumpers.
Optimal conditions at take-off in the long jump are influenced by what happens on the last few steps to touch-down onto the take-off board. Lowering of the center of mass on the second-to-last step, achieved by lengthening the second-to-last step length (8), and/or landing onto the last step with a more flexed knee (13), is important because it provides optimal body posture for the start of the TO phase and avoids a high negative vertical velocity at touch-down. For amputee athletes, particularly those with a transfemoral amputation, the fact that the prosthetic knee needs to be locked to support body weight during stance/load bearing means they have difficulties in lowering their center of mass on prosthetic limb-support steps.
It has also been reported that there are gender differences in how the center of mass is lowered. One study (6) showed that female able-bodied long jumpers had a lower center of mass than males on the approach run but then lowered their center-of-mass height less on the last step and at touch-down. Differences in eccentric leg strength were suggested as one possible factor for these gender differences. Another study (1) observed that there also tended to be gender differences for transtibial athletes in manipulating the center-of-mass height during the last few steps before touch-down. Whereas male transtibial athletes lowered their center of mass during the second-to-last step, females raised theirs. However, because there was such a small sample of female jumpers, the author concluded that neither the underlying reasons nor the effects on performance for these differences could be determined.
Hence, it is evident that kinematic factors related to the approach and take-off phases that underlie technique for a successful flight phase are not known for female amputee long jumpers and may differ from the techniques used by their male counterparts. Therefore, the purpose of this study was to investigate whether female lower-limb amputees conform to the established long-jump model and to compare the kinematics of the approach and take-off phases for elite female transfemoral and transtibial amputee long jumpers.
A biomechanical analysis of the long-jump technique used by female transfemoral (F42) and transtibial (F44) athletes during the finals of the 2004 Paralympic Games was performed. All those athletes who used a lower-limb prosthesis during the competition were included in the analysis, which included eight transfemoral and nine transtibial amputee athletes.
Apparatus and procedure.
Approach distance and speed during the run-up of each jump was recorded (100 Hz) using a laser Doppler device (LDM 300 C Sport, Jenoptik Laser, Jena, Germany). The device was positioned approximately 60 m behind the long-jump pit, elevated approximately 20 m with the origin (0 m) set to the take-off board at torso height. Perspective error was automatically corrected for before recording using the option in the Doppler device sampling software (DAS3E v3.9, Jenoptik Laser, Jena, Germany). The values of the actual measured distance (horizontal) and elevation (vertical) from point of origin (0 m), obtained manually (Fig. 1), were typed into the software. The laser was then targeted onto the point of origin and the software automatically corrected itself for the perspective error using the Pythagorean theorum. During the approach, the device was targeted on the torso of the athlete. For each successful jump, speed data at the point of take-off (0 m) were obtained and smoothed (13-point moving average) using Distance Evaluation Sport software (DAS3E v3.9).
Two digital video cameras (Sony, model DCR-TRV33E), recording sagittal plane movements at 50 Hz, were placed so that the third-to-last step on the approach run to take-off from the board were visible. Before competition, a 3D, 18-point calibration frame was filmed at several points along the approach run. During the competition, the cameras were left running and the athletes' jumps were recorded. Official distance was documented for every jump, and "effective distance" was later calculated from official distance plus the toe-to-board distance, which was measured from the sagittal plane video one frame before take-off.
The best jump (greatest official distance) for each athlete was then selected for detailed kinematic analysis. The video data were digitized using eHuman digitizing software (HMA Technology, Inc, Ontario, Canada). The 2D coordinates for each stride were reconstructed from the sagittal plane data and analyzed using a nine-segment biomechanical model defined by 18 points. These were C7 and top of the head, bilateral wrist, elbow, shoulder, hip, knee- and ankle-joint centers, and the toe and heel. Anthropomentric data were taken from Dempster (3) for adult females. Adjustments to the anthropometric data to account for the prosthetic limb were made in the same way as described previously (9). The data were smoothed using a Butterworth fourth-order digital filter and a cut-off frequency of 7 Hz. Because measurements of the athletes' heights were not available, each athlete's estimated height was calculated as the sum of the length of her individual segments determined from the coordinates of the digitized data (6). These segments, which included (i) bottom of the intact-limb foot to the ankle, (ii) intact-limb ankle to knee, (iii) intact-limb knee to hip, (iv) hip to shoulder, and (v) neck to top of head, were summed for each athlete to obtain an estimation of her individual height. Coordinate data for the prosthetic limb were not included because athletes can set their prosthesis length slightly longer or shorter than their intact limb depending on personal preference.
The critical events of touch-down (TD) and take-off (TO) for the third-to-last step (3LS), second-to-last step (2LS), last step (LS), touch-down on the take-off board (TDboard), the point of maximum knee flexion while in contact with the board (MKFboard), and take-off (TOboard) were identified for each athlete. The point of TD was taken as the first frame in which the foot was clearly seen to be in contact with the ground, and TO was taken as the first frame in which the foot was seen to leave the ground. Kinematic variables were then calculated for TD and TO on each step and from TDboard to TOboard. These included center-of-mass height (%HCM), horizontal and vertical velocity at each step, hip angle, knee angle, and leg angle at TD. The height of the center of mass obtained for each athlete was normalized to her estimated height. The hip and knee angles were defined as the included angles between shoulder, hip and knee, and hip, knee, and ankle, respectively. The leg angle was defined as the angle made by the line joining the center of mass and the ankle to the vertical (Fig. 2).
The maximum possible error associated with sampling at 50 Hz for each key variable was determined in the following way. Data for three of the athletes were analyzed for two consectutive frames at TDboard, where TDboard was clearly occurring either between these two frames, or at one of these two frames. The same was done for TOboard. The maximum difference in each variable between the two frames for both TD and TO was taken to be the maximum possible error attributable to sampling frequency. Table 1 shows the results of this error analysis. Whereas it is generally agreed that 50 Hz is the limit of an acceptable sample frequency for motion analysis in elite sport, these results show that the maximum errors attributable to sampling frequency are within the standard deviation of the transfemoral and transtibial groups (Figs. 5-7). For example, the maximum error at TD would result in a 1-cm increase in HCM, a 0.1-m·s−1 increase in horizontal velocity, a 7° increase in hip angle, and a 4° increase in knee angle (Table 1). There was a quite large increase in vertical velocity at TD, particularly for the two transfemoral amputees, but this would only serve to make the difference between the groups larger. At TO, there was a maximum increase of 4 cm in HCM, a 0.1-m·s−1 increase in both horizontal and vertical velocity, a 10° increase in hip angle, and a 3° increase in knee angle. Because the errors attributable to using a 50-Hz sampling frequency are within the group standard deviations for this population, these data are deemed valid.
Linear regression was performed to determine the relationship between TO speed (recorded during the approach run) and effective distance jumped. A Shapiro-Wilks test was used to test the normality of the kinematic data. A Mann-Whitney U-test established differences between all transfemoral and transtibial athletes on each step, and the Wilcoxon test was used to indicate differences between steps for each group. Spearman's rank correlation coefficient was used to establish relationships between the variables. The significance level used was alpha = 0.05. Bonferroni adjustments to the alpha level were not made because although this correction reduces the likelihood of making a Type I error, it increases the probability of making a Type II error, putting one in danger of not finding any differences in technique where there actually are, particularly when sample sizes are small. To avoid these errors, key variables to determine technique were specifically chosen, and analyses between variables were limited to avoid unnecessary multiple tests.
Table 2 indicates distance jumped and choice of TO leg for each athlete. Because it has previously been shown that choice of TO leg does not affect joint kinematics for transtibial athletes (10), and only one transfemoral athlete took off from her prosthetic limb, the results are presented in two groups.
Figures 3 and 4 illustrate the relationship between TO speed and effective distance jumped. For the transfemoral athletes, there was no relationship between these variables (r = 0.100). For the transtibial athletes, there was a significant (P < 0.05) relationship (r = 0.668). Thus, for an increase in approach speed, female transtibial athletes increased jump distance, whereas female transfemoral athletes appear to be relying on a mechanism(s) other than their approach speed to influence distance jumped.
The transfemoral athletes had a slower horizontal velocity of the center of mass (P < 0.05) on all approach steps than the transtibial athletes (Fig. 5). The transfemoral athletes reduced horizontal velocity (P < 0.05) between 2LSTD and LSTD, then increased (P < 0.05) LSTD to TDboard. All athletes reduced horizontal velocity (P < 0.05) between TDboard and TOboard.
Vertical velocity of the center of mass was similar for transfemoral and transtibial athletes on the last three approach steps, being negative on each step TD and positive on each TO (Fig. 6). At TDboard, the transfemoral athletes had a significantly greater (P < 0.05) negative vertical velocity than the transtibial athletes. However, at TOboard, vertical velocity was almost the same between the two groups (Fig. 6).
For the amount of horizontal velocity lost and vertical velocity gained from TDboard to the end of the TOboard phase, the transtibial athletes lost more (P < 0.05) horizontal velocity from TDboard to MKFboard (0.56 m·s-1) than the transfemoral athletes (0.27 m·s-1). No significant difference existed between groups for vertical velocity gained during any part of the TOboard phase.
The relative center-of-mass height (%HCM) was consistently higher for the transfemoral compared with the transtibial athletes during the last three approach steps (Fig. 7). At 2LSTO, there was a significant difference in %HCM (P < 0.05) between the two groups, which may reflect an adjustment the transfemoral athletes made in preparation for TDboard. However, the transfemoral athletes lowered their center of mass much more at TDboard from their LS (P < 0.05) than the transtibial athletes. Consequently, %HCM was slightly lower for transfemoral athletes than for transtibial athletes from TDboard to MKFboard. However, at TOboard, %HCM was similar between the two groups.
On the 3LS and LS, transfemoral athletes touched down with less (P < 0.05) hip and knee flexion (i.e., more upright) than transtibial athletes (Table 3). A greater hip range of motion (P < 0.05) was observed for transfemoral athletes on these steps (Table 4), which may indicate that these athletes compensate for limited use of the knee joint by relying more on hip movement. At TDboard, the two groups exhibited exactly the same mean hip angle, although transtibial athletes had a more flexed knee (P < 0.05). However, by TOboard, the transfemoral athletes had extended the hip much more (P < 0.05). Consequently, hip range of motion was greater (P < 0.05) for transfemoral athletes from TDboard to TOboard. Because hip range of motion was the same between the two groups from TDboard to MKFboard, the results show that the transfemoral athletes increase hip range of motion the most from MKFboard to TOboard (Table 4).
Transfemoral athletes exhibited a smaller (P < 0.05) leg angle (more vertical) on the 2LS and TDboard than transtibial athletes. This, together with the greater hip and knee angles on the 2LS, puts them in a much more "upright" position. At TDboard, the similarity in hip and knee angles but the difference in leg angle may indicate that whereas the transfemoral athletes touch down in an upright position, the transtibial athletes may be leaning slightly backward.
Considering the kinematic differences step by step (Table 3), transfemoral athletes increased (P < 0.05) hip and knee angle on LSTD, then decreased (P < 0.05) again at TDboard. They also had a much more horizontal leg at TDboard (P < 0.05) compared with at their previous steps. The more flexed hip and knee and more horizontal leg at TDboard resulted in a lowering of the %HCM. The transtibial athletes, in contrast, did not significantly alter their hip and knee angles during the last three approach steps. However, from TDboard to MKFboard, they increased hip angle (P < 0.05) and decreased knee angle (P < 0.05) (Table 3).
All of the female amputee athletes in this study had a slower approach speed and jumped less far than elite female able-bodied long jumpers (7). However, the transtibial athletes were slightly faster and jumped slightly farther than the results reported for four female transtibial athletes competing in an international competition in 1998 (1). The relationship between approach speed at TO and distance jumped has previously been found to be strong for female able-bodied athletes (7), male able-bodied athletes (8), male transtibial amputees (r = 0.85, 0.90), and male transfemoral amputees (r = 0.83, 0.85) (2,9). Although there was a weaker relationship for female transtibial amputees in this study (r = 0.668), the complete lack of any relationship for female transfemoral amputees indicates that either some of the jumps included in the analysis were poor because of errors made by the athletes on their approaches (taking into account only the best jumps, the relationship was still nonsignificant), or that what these athletes are doing during the TOboard phase is not allowing them to take full advantage of the horizontal velocity developed during the approach run. Thus, the results indicate that the long-jump model that holds true for able-bodied athletes, male amputee athletes, and female transtibial amputee athletes does not hold for female transfemoral athletes.
The optimal position of the body at TDboard is dependent on what the athlete does in the last few approach steps. The higher center-of-mass position for female transfemoral compared with female transtibial athletes on the last few steps has also been previously reported for male amputee athletes (10) and may be attributable to the fact that the prosthetic knee mechanism cannot flex in stance while load bearing. What is different in the present study, however, is that the female transfemoral athletes lowered their %HCM so much on their LS that they were lower than the female transtibial athletes from TDboard to MKFboard. One study (6) reported that although able-bodied female athletes lowered their %HCM on LS and TDboard, they did not lower it as much as the males did. This was attributed to the male athletes being stronger eccentrically and more able to control and benefit from greater hip, knee, and ankle flexion on the TD leg. If the center of mass is lowered too much for the amount of eccentric leg strength, the leg will buckle, resulting in a less than optimal jump (2). Whereas the female transfemoral athletes touched down onto the TO board with a hip and knee angle similar to that of the female transtibial athletes, the transition from a much more upright, much higher center-of-mass position on the previous step means that the transfemoral athletes would have to have greater eccentric leg strength to lower themselves by such an amount and effectively sustain the forces acting on the leg at TD. This may explain why these female transfemoral athletes exhibited no relationship between TOboard speed and distance jumped and, thus, did not gain the expected increase in distance jumped for an increase in approach speed, unlike male transfemoral athletes in previous studies (9,11).
The step on which the center of mass is lowered influences TO parameters. Lowering the center of mass on the LS, as seen by these female transfemoral athletes, results in a disadvantageous, high negative vertical velocity in the beginning of the TO phase (Fig. 4), which needs to be reversed before TOboard. By lowering the center of mass on the 2LS, either by taking a longer step or by increasing knee flexion and staying low into the LS, able-bodied long jumpers avoid the problem of touching down onto the TO board with a high negative vertical velocity (6). However, because of the constraints of the prosthetic knee mechanism, transfemoral athletes are unable to do this.
One study (7) reported vertical velocities of −0.03 m·s−1 at TDboard, 1.98 m·s−1 at MKFboard, and 3.05 m·s−1 at TOboard for female able-bodied long jumpers. The female amputee athletes in this study had a greater negative vertical velocity at TDboard (−0.18 and −0.39 m·s−1 for transtibial and transfemoral athletes, respectively) and were not able to produce the same high vertical velocity at MKFboard (0.85 and 0.79 m·s−1) or TOboard (1.83 and 1.81 m·s−1), limiting their jump performance. The inability to lower the center of mass on the optimal stride for the transfemoral athletes and, to a lesser extent, for the transtibial athletes, may be attributable to problems with the prosthetic limb. For the transfemoral athletes, this problem is greater because they are unable to actively flex both their prosthetic knee and ankle. In addition, transfemoral athletes seem to need to raise their center of mass even higher on 2LSTO, possibly for the prosthetic limb to clear the ground during the swing phase.
The consistently higher center of mass of the transfemoral athletes on the last few steps may be a result of the more extended hip and knee at TD on every step. Touching down onto the prosthetic limb, the leg needs to be quite far under the body for support. This forces the athlete to be in a quite upright position. Because of the absence of prosthetic knee flexion during swing, the athletes also need to touch down onto the intact limb in a fairly upright position to clear the foot from the ground.
Both the female transfemoral and transtibial athletes touched down onto the TO board with a more flexed knee (154 and 149°, respectively) than that reported for able-bodied female long jumpers (160°) (7). This has also been observed for male amputee athletes in comparison with male able-bodied athletes (9). Elite male transfemoral long jumpers have been found to exhibit a more flexed hip position at TDboard to increase hip range of motion as a compensatory mechanism for having to touch down in a more upright position (9). Thus, the female transfemoral athletes are using a similar technique, which was seen in this study by the greater hip range of motion from TD to TO compared with the female transtibial athletes. It may be that trying to rely more on hip strength and increasing hip range of motion during the TO phase and less on lowering the center of mass excessively at TDboard could be an advantage for female transfemoral long jumpers.
In summary, female transtibial athletes conformed to the long-jump model, although adaptations to this technique were displayed. However, these adaptations did not allow these athletes to make as effective use of the horizontal approach speed as able-bodied and male amputee athletes. Female transfemoral athletes, however, exhibited no relationship between TOboard speed and distance jumped, which may be attributable to their excessive lowering of their %HCM at TD onto the TO board, causing the leg to buckle. It is recommended that coaches and athletes proceed with caution when trying to replicate techniques used by able-bodied athletes because adaptations to the constraints a prosthesis should be considered.
This study was funded by CIF (Centre for Sport Research, Stockholm, Sweden) and The Swedish School of Sport and Health Sciences, GIH, Stockholm. The authors would like to express their appreciation to Morgan Nolan and Kjartan Halvorsen.
1. Ciapponi, T. A. A kinematic analysis of techniques used by elite transtibial and transfemoral amputee long jump athletes. Unpublished Ph.D. Thesis. University of Georgia, Athens, GA, 2000.
2. Dapena, J. Basic and applied research in the biomechanics of high jumping. Med. Sport Sci.
3. Dempster, W. T. Space requirements of the seated operator. WADC technical report, Wright-Patterson Air Force Base, OH, 1955, pp. 55-159.
4. Hay, J. G. Citius, Altius, Longius (Faster, Higher, Longer): The biomechanics of jumping for distance. J. Biomech.
26 (Suppl. 1):7-26, 1993.
5. Hay, J. G., J. A.Miller Jr., and R. W. Canterna. The techniques of elite male long jumpers. J. Biomech.
6. Hay, J. G., and H. Nohara. Techniques used by elite long jumpers in preparation for take-off
. J. Biomech.
7. Lees, A., N. Fowler, and D. Derby. A biomechanical analysis of the last stride, touch-down and take-off
characteristics of the women's long jump. J. Sports Sci.
8. Lees, A., P. Graham-Smith, and N. Fowler. A biomechanical analysis of the last stride, touch-down, and take-off
characteristics of the men's long jump. J. Appl. Biomech.
9. Nolan, L., and A. Lees. The kinematic characteristics of above and below knee amputee long jumpers. Ergonomics
10. Nolan, L., and A. Lees. Prosthetic limb versus intact limb take-off
in the amputee long jump. In: Proceedings of The XXIII International Symposium on Biomechanics in Sports.
, Beijing, China, August 22-27, 2005, pp. 305-308.
11. Patritti, B. L., K. J. Simpson, and L. Nolan. Approach
velocity profiles of elite male and female lower-limb amputee long jumpers. In: Proceedings of the XXth Congress of the International Society of Biomechanics
, Cleveland, OH, 2005, p. 823.
12. Simpson, K. J., S. L.Williams, P. DelRey, T. M. Ciapponi, and H. L.Wen. Locomotor characteristics exhibited during paralympic long jump competitions of classifications "below-" and "above-knee amputee". Int. J Appl. Sport Sci.
13. Tidow, G. Models for teaching techniques and assessing movements in athletics. The long jump. Track Technique
Keywords:©2006The American College of Sports Medicine
TAKE-OFF; APPROACH; DISABLED; SPORT