In basketball, a running single leg jump (RSJ) is frequently used in various competitive activities such as rebounding, blocked shots and lay-up shots at a higher position. Hence, the ability of the RSJ is a determinant for achieving high competitive performances in this sport. It is known that basketball players have a dominant takeoff leg (DL), which is preferably used to perform the RSJ (Schiltz et al., 2009). For basketball players, however, it is critical to perform the RSJ at the same level regardless of the DL and non-dominant legs (NDL), because they must offend and defend from various directions and situations relative to the goal. In addition, Brumitt et al. (2013) and Schiltz et al. (2009) argued that limb asymmetry in performance over 10% increases the occurrence of sport-related injuries. For basketball players, lower extremities are the most frequently injured body area (Borowski et al., 2008; Decker et al., 2003; Ellapen et al., 2012; Messina et al., 1999; Nelson et al., 2007) because they frequently use each leg to jump, pivot or change directions, and are likely to overuse the takeoff leg as it enables them to jump higher. Thus, to reduce any asymmetry in the performance of the RSJ is essential, not only to achieve high performance, but also to prevent injury to the lower limbs.

Many studies have already investigated asymmetry between limbs in performance during different single leg jumps, which are started from a standing position (Hewit et al., 2012; Maulder and Cronin, 2005; Meylan ets al., 2010; Miyaguchi and Demura, 2010; de Ruiter et al., 2010; Schiltz et al., 2009; de Swearingen et al., 2011). Some studies (Meylan ets al., 2010; Miyaguchi and Demura, 2010; Swearingen et al., 2011) showed that the performance of a single leg vertical jump from standing was significantly higher in the DL than in the NDL. Meylan et al. (2010) also observed a similar phenomenon in the horizontal and lateral counter movement jumps. In addition, Schiltz et al. (2009) showed that professional basketball players jumped significantly higher with the DL than the NDL (12%) in drop jump. These studies focused on jumps from a standing position, and to our knowledge, no studies have tried to examine the asymmetry of performance in the RSJ.

Many studies have already investigated the kinematics during RSJ through comparing high jumping players with low or correlating the jump height with jump variables in basketball and jump events of track and field (Greig and Yeadon, 2000; Ritzdorf, 2009; Stefanyshyn and Nigg, 1998). Stefanyshyn and Nigg (1998) have reported that to exert a greater force by ankle angle compared with knee and hip joints is important to jump high in RSJ. Moreover, it has been shown that high jumpers kept knee joint to more extended position to jump high at takeoff phase (Greig and Yeadon, 2000). Considering these findings, it may be assumed that the asymmetry of performance in RSJ is attributable to that of the kinematics of takeoff leg.

The present study aimed to clarify the differences between the DL and NDL in the behavior of the lower extremity during RSJ in collegiate basketball players. To this end, we compared the kinematics of the lower limbs in the takeoff leg during the RSJ between the DL and NDL and examined how the asymmetry of the jump height can be associated to that in the kinematics of the takeoff legs. We hypothesized that the asymmetries of jump height in the RSJ can be associated to that in the kinematics of the takeoff leg, notably in the ankle joint.

Twenty-seven healthy male collegiate basketball players voluntarily participated in this study. Their means and standard deviations (SDs) of age, body height, and body mass were 21.1 ± 2.4 yr, 1.76 ± 0.05 m, and 71.4 ± 7.7 kg, respectively. Fifteen of the subjects were guard players, nine forward players and three center forward players except pure center playing near the rim. The subjects had experienced organized competitive basketball training for at least eight years and participated in domestic or inter-college competitive meetings. They were in good physical condition and without any history of injury to the lower limbs within the last year. On the basis of questionnaires, the DL was defined as the one used preferably for jumping with a single leg, as described in prior studies (Schiltz et al., 2009). In this study, all subjects could jump higher with the DL than the NDL during RSJ. The DL for 18 of the subjects was the left; therefor, for the remaining subjects was the right. This study was approved by the Ethics Committee of the National Institute of Fitness and Sports in Kanoya and was consistent with institutional ethical requirements for human experimentation in accordance with the Declaration of Helsinki. Prior to the measurement session, all subjects were fully informed about the procedures and possible risks involved as well as the purpose of the study. Written informed consent was obtained from all subjects.

The participants were asked to perform a maximal RSJ that was similar to a condition in which basketball players rebound a ball close to the rim. The subjects approached for 6 m in which the distance was similar 3-point line when outside players started playing in the area at their own speed and jumped maximally in order to touch by hand a target set to the end of a ceiling 3.05 m in height. The experimental test was carried out without the ball to eliminate various constraint conditions. This allowed the participants to fully concentrate on performing the jump task itself. Execution of the RSJ was firstly demonstrated by an experimenter, and then the subjects were asked to practice the task to familiarize themselves with the testing procedure. In the measurements, the subjects were asked to jump four times with each takeoff leg alternated, with a sufficient rest interval which did not influence next jump performance. We calculated a symmetry index of jump height (symmetry index_JH) and index of each kinematics data as {2 × (DL – NDL) / (DL + NDL)} × 100 (Wong et al., 2007). The symmetry index_JH was 6.3 ± 4.8 % (means ± SDs). In another experiment using the subjects examined here, we determined the mechanical power developed during a single leg squat vertical jump, which was without preparatory counter movements and with their arms akimbo, and confirmed that there was no significant difference in mechanical power between the two legs.

The motions of the segments and contact time from touchdown to takeoff in the RSJ were recorded in three-dimensions using a 12 Raptor camera infrared motion analysis system (MAC 3D system, Motion Analysis, Corp., USA) at a sampling frequency of 500 Hz. The data was stored on a personal computer system through an A/D converter. Experimental 3-D error of capturing marks was less than 1 mm in all trials in the current study. Twenty-six reflective markers of 1.4 cm diameter were placed at specific landmarks of the body (Figure 1). Absolute coordinates of these markers were chosen to represent the joints of lower limb and to calculate the center of gravity (CG). The horizontal distance between the takeoff place and the vertical plane containing the target was adjusted for each participant so that they could jump with maximal effort. A zero-lag, fourth-order, low-pass Butterworth digital filter with a cut-off frequency of 12 Hz was chosen as the best presentation of the raw positional data of the markers after a visual inspection of several frequencies. The cut-off frequency was selected in accordance with prior studies (Ae et al., 2008; Hay et al., 1999; Wang, 2011).

Among the four trials carried out by the subjects, the highest jump trial in each of the DL or NDL was selected for further analysis. In this study, the jump height was calculated as the vertical height traveled by the marker attached to the sacral bone in the apex during the RSJ from a standing position. Joint angles and angular velocities of the lower extremities in the takeoff leg during takeoff phase, contact time (CT) to ground during the jump and maximal velocity in the running approach were calculated from the 3-D data (Figure 2). The CT was defined as from the landing of a heel to taking off from a toe, including the CG descent time (CT_des) and the CG ascent time (CT_asc) based on minimal vertical displacement of the CG (CG_zmin). The angle and angular velocity of the takeoff leg was analyzed for peak value in maximal (max), minimal (min), CG_zmin and average value of CG_des (Ave_des) and of CG_asc (Ave_asc).

Descriptive data are presented as means ± SDs. A Stu-dent’s paired t-test was used to test the significance of the differences between the two legs in the measured variables. Using Pearson’s product moment correlation analysis, linear correlation coefficients between the symmetry index_JH and indices of kinematics data were calculated. The level of significance was set at p < 0.05. All data were analyzed using SPSS software (SPSS statistics 22, IBM, Japan).

There was no significant difference in running velocity during approach to the RSJ between the trials of the DL (7.9 ± 0.5 m·s^–1) and NDL (7.8 ± 0.4 m·s^–1, Table 1). However, the jump height was significantly higher (p < 0.001) in the DL than in the NDL (Table 1). The CT for the DL was significantly shorter (p < 0.05) than that for the NDL (Table 1). The symmetry index_JH was 6.3 ± 4.8 %, and it was over 11.1% (mean + 1 SDs) in 6 subjects, from 8.5 to 11.0% (mean + 0.5 SDs) in 5 subjects, from 1.5 to 3.9% (mean - 0.5 SDs) in 6 subjects, under 1.5% (mean - 1 SDs) in 5 subjects, and from 4.0 to 5.9% in the remainder (n = 5).

Table 2 summarizes descriptive data on joint angles for the DL and NDL. The min, Ave_des, Ave_asc and CG_zmin values of ankle angles were significantly larger in the DL than in the NDL (p < 0.05, 0.01). The min, Ave_asc and CG_zmin of knee angles were significantly larger in the DL than in the NDL (p < 0.05).

Table 3 summarizes descriptive data on joint angular velocity for the two legs. The Ave_asc of ankle joint angular velocity was significantly greater in the DL than in the NDL, and that of knee joint angular velocity was vice versa (p < 0.05).

The symmetry index_JH was not significantly correlated to those of running velocity and CT (Table 4). On the other hand, the symmetry index_JH was positively correlated to the index of each of the min, Ave_des and Ave_asc values of the angle (p < 0.05, 0.01) and of the Ave_asc value of the angular velocity (p < 0.05) in the ankle joint (Table 5).

The jump height and knee and ankle joint kinematics significantly differed between the DL and NDL. Furthermore, the symmetry index_JH was positively correlated with the index of each of the angle values, and angularvelocity, in only the ankle. These results support our hypothesis that the asymmetry of the RSJ performance can be associated with that of the behavior of the takeoff legs, notably in the ankle joint.

For the knee joint, the min, Ave_asc and CG_zmin values of the angle were larger in the DL than in the NDL, although the Ave_asc of the angular velocity was greater in the NDL. This implies that, in the takeoff phase, the knee joint in the DL was in a more extended position as compared to the NDL. The jump height for high jumpers has been shown to be maximized when the knee joint of the takeoff leg is maintained in an extended position at touchdown (Greig and Yeadon, 2000). Knee kinematics at landing in a RSJ are an important factor affecting the ground reaction force (GRF) in vertical jumps (Wang, 2011). In jumps including a running or drop landing task, the decreased flexion of the maximal hip and knee angles at landing are linked to the increased vertical GRF (Decker et al., 2003; Yu et al., 2006). Taking these findings into account together with the current results, it may be assumed that the DL can produce a vertical impulse to the ground and effectively transmit the vertical GRF to the upper limb by keeping the leg straight (i.e., stiffen the knee joint) in order to jump high. However, it should be noted that, despite the significant differences between the two legs for the kinematics of the knee joint, their symmetry indices were not associated with the symmetry index_JH. This suggests that the observed bilateral differences in the kinematics of the knee joint are not a determinant factor for the symmetry index_JH in RSJ.

For the ankle joint, the min, Ave_des, Ave_asc and CG_zmin values of the angle and the Ave_asc of angular velocity were also greater in the DL than in the NDL. This indicates that, in the takeoff phase, the ankle joint in the DL maintained a more plantar flexed position than that in the NDL, with a greater angular velocity in the push off phase. This may be linked to the result that the CT was significantly shorter in the DL than in the NDL, while the run-up velocity was similar in both legs because it is known that the stiffness of the ankle joint plays an important role in jumping high and taking off in a short time in the explosive jumps and drop jumps (about 0.2 sec, Yoon et al., 2007). Also, Stefanyshyn and Nigg (1998) reported that the ankle joint, as compared to the knee and hip, was the largest energy absorber and generator in RSJ and long jump. In the rebound jumps, the stiffness of the ankle during the eccentric phase is positively associated to the ankle joint torque at the turning point from an eccentric to a concentric activity (Yoon et al., 2007). The ankle torque also showed significant positive correlation with jump height (Yoon et al., 2007). Farley and Morgenroth (1999) reported that, during hopping in place, the stiffness of the ankle was 1.9 times greater while that of the knee was 1.7 times greater at the maximal height than in the preferred height jump (Farley and Morgenroth, 1999). Considering these findings, the stiffness of the ankle compared to the knee clearly plays an important role in explosive jumps including RSJ and standing jump. In the current results, therefore, the observed differences between the two legs in the angle and angular velocity of the ankle joint (i.e., a straightness of the ankle and explosive bounding from CG_zmin) suggest that the landing energy absorbed and the energy used for jumping are smaller in the NDL (vs. the DL), and consequently, this may have resulted in the difference in jump height for each leg. At the same time, the current result that the symmetry indices of min, Ave_des and Ave_asc for ankle angle, and of Ave_asc for the joint angular velocity, were significantly associated with the symmetry index_JH suggests that the bilateral difference in the behavior of the joint directly influences the magnitude of the difference in the performance of RSJ.

It is known that basketball players have a leg which is used preferably for performing a RSJ (Schiltz et al., 2009). The DL for the players develops in the process of training to a professional level (Theoharopoulos and Tsitskaris, 2000). However, a limb asymmetry of over 10% can be considered a risk factor for inducing sport-related injuries (Brumitt et al., 2013; Schiltz et al., 2009). Considering this, it may be assumed that players with a large asymmetry in their performance have a greater risk of sport-related injuries, but this would be limited to regular players. In contrast, in the current study, only 3 of the 11 players who had an asymmetry of the symmetry index_JH over 8.5% (+ 0.5 SDs) were regular players who were starters or replaced the starters in competitive games. All 11 players who scored less than 3.9% (- 0.5 SDs) were regular, while 8 players were non-regulars of those with asymmetry over 8.5%. Finally no one was less than 3.9%. These results refute the previous report (Theoharopoulos and Tsitskaris, 2000) cited above and indicate that, at least at the collegiate level, basketball players with a large asymmetry in their jump performances would be mainly judged as non-regulars.

The current results indicate that, for collegiate basketball players, the asymmetry in the height of a RSJ can be attributed to the joint kinematics of the ankle during the takeoff phase, which may be associated with the ability to effectively transmit run-up velocity to jump height. Additionally, basketball players who have a large asymmetry of the RSJ at the collegiate level could be assessed as non-regulars judging by the magnitude of asymmetry rather than jump height.