Changes in Lower Extremity Kinematics and Temporal Parameters of Adolescent Baseball Pitchers During an Extended Pitching Bout

Participants

Twelve male adolescent pitchers (mean age, 14.9 ± 0.7 years; mean height, 1.74 ± 0.07 m; mean weight, 72.8 ± 13.5 kg) were recruited through local youth baseball teams and public advertisements. To be eligible for recruitment, athletes must have had pitched within the previous month but not within 5 days before testing. Participants were excluded if they were previously diagnosed with shoulder pain or dysfunction by a physical therapist or had a history of injuries or surgery to the shoulder of the throwing arm. Nine participants were right-hand dominant, and 3 were left-hand dominant. Participants had a reported mean of 5.2 ± 2.1 years of pitching experience based on a self-reported survey. Participants and their parent/guardian provided informed written assent and consent, respectively. The study was approved by an institutional review board.

Marker Placement

Participants wore their own shorts and athletic shoes. Retroreflective markers were attached to the following bony landmarks by a single observer: the acromioclavicular joint of the nonthrowing shoulder, manubrium and xiphoid process of the sternum, C7 and T8 spinous processes, sacrum (midpoint of L5 and S1 spinous processes), left posterior superior iliac spine, and bilateral anterior superior iliac spines (ASISs) (Figure 2). Markers were also attached to the outer surface of the shoes over the approximated bases of the first and fifth metatarsals, and 4 markers were attached to the posterior surface of the heel of the shoe. An adjustable but rigid 4-marker cluster was attached to the scapular spine of the throwing arm. ¹² Additional 4-marker clusters were placed on the proximal end of the ulna inferior to the olecranon process, the dorsal side of the distal forearm, and the dorsal surface of the throwing hand and bilaterally on the lateral side of the thigh and shank. The rigid clusters were secured using self-adhesive bandages (Coban; 3M). The aforementioned rigid marker clusters and single tracking markers were not removed during the pitching protocol. Lower extremity calibration markers were attached bilaterally on the greater trochanter, medial and lateral femoral condyles, and ankle malleoli and over the second toe. The calibration markers were removed after participant-specific calibration trials and before the pitching trials. Reflective tape was also fixed to the sides of the baseballs to help identify the ball release event.

OPEN IN VIEWER

Figure 2.

Anterior and posterior views of the marker set for a right-handed pitcher (scapula, ulna, radius, and hand clusters were on the left arm for left-handed pitchers). The rigid clusters of markers (solid black circles) and single markers (gray circles) were used as tracking markers. White circles identify the calibration markers. Markers on the trunk and upper extremities were only used for event identification (ie, ball release and maximum shoulder internal rotation).

Data Collection

A 12-camera optical motion capture system (Motion Analysis Corp) was used to collect and track 3-dimensional marker trajectories at a sampling rate of 250 frames per second. Before marker calibration, participants completed their own warm-up routine including stretching, followed by 10 warm-up pitches, while wearing the full marker set. A static calibration trial was performed while participants stood upright and looked straight ahead. Using a previously developed regression definition, ¹⁷ the positions of the hip joint centers were calculated relative to the pelvic coordinate system using the inter-ASIS distance. Knee and ankle joint centers were created virtually using the midpoint of the femoral condyles and malleoli markers.

Participants completed 6 sets of 15 fastball pitches (ie, 90 pitches) with 5-minute rest periods between sets. ⁶ Participants pitched from a sloped 10 inch–high artificial pitching mound in front of a pitching target. The target was placed behind a plastic home plate set at the high school regulation distance of 18.4 m (60.5 ft) from the mound rubber. Ball speed was obtained using a radar gun (Sports Radar DT200; Sports Radar Ltd), which was placed near the target. Pitching accuracy was assessed by a rater referencing a rectangular strike zone (0.38 m wide × 0.64 m tall; 0.46 m from the ground) on the target. A “strike” was a pitch that hit the pitching target within the strike zone. Inaccurate or “ball” pitches were divided into 2 subcategories: “ball 1” and “ball 2.” A ball 1 pitch hit the pitching target but outside the strike zone (ie, a 0.54-m margin of error around the strike zone). A ball 2 was a pitch that missed the pitching target structure completely: a “wild pitch.” The number of strike, ball 1, and ball 2 pitches counted in the baseline and final sets were expressed as a percentage of the 15 pitches thrown during the respective sets.

Data Processing and Analysis

Motion data were processed and analyzed in Visual3D (C-Motion Inc). Marker trajectories were smoothed using a fourth-order low-pass Butterworth filter with a 12-Hz cut-off frequency for the lower extremity and pelvis markers. The cut-off was based on the highest average frequency at 95% cumulative power spectral density of the X, Y, and Z trajectories of a single participant’s pelvis and lower extremity marker data. The following events during the pitching motion were identified: (1) the balance position, identified as the instant when the leading knee reached its maximum height ⁹ ; (2) foot contact, identified as when the velocity of the leading foot fell below 0.5 m·s⁻¹; (3) ball release, identified as the first frame in which the ball had left the hand; and (4) maximum shoulder internal rotation. The pitching cycle was defined as the balance position (0%) through maximum internal rotation of the shoulder (100%). The pitching cycle included the (1) stride phase (balance position to foot contact), (2) acceleration phase (foot contact to ball release), and (3) deceleration phase (ball release to maximum shoulder internal rotation) (Figure 1).^1,24

Stride length was calculated as the displacement between the ankle joint centers of the trailing limb at the balance position and the leading limb at foot contact. ² Stride length was normalized to body height. Foot progression was calculated as the angle between the long axis of the foot relative to the throwing direction, that is, the direction of home plate. The foot progression angle was 0° when the foot was facing the pitching net (ie, center of home plate). A positive foot progression angle indicated that the foot was medially rotated (toed-in). Similarly, pelvis orientation in the transverse plane was calculated relative to the throwing direction. The pelvis orientation was 0° when the pelvis was facing toward home plate. A positive value indicated that the pelvis experienced transverse rotation of the dominant side in the global coordinate system (ie, counterclockwise, or right transverse rotation of the pelvis for a right-handed pitcher). Foot progression and pelvis orientation were calculated at the balance position, foot contact, and ball release events.

Angular excursions (deg) and angular velocities (deg·s⁻¹) were calculated for the hip (sagittal, frontal, transverse planes), knee (sagittal plane), and ankle (sagittal plane) of the leading leg at each event (ie, balance position, foot contact, and ball release). Angular excursions of the trailing leg at the hip, knee, and ankle were calculated at the balance position. Maximum angular excursions were calculated at the hip, knee, and ankle during the stride phase for both legs. Maximum angular velocities of the trailing leg at the hip, knee, and ankle during the stride phase as well as the maximum knee extension angular velocity of the leading leg during the acceleration phase were calculated.

The duration of the pitching cycle was defined as the time between the balance position and maximum shoulder internal rotation (in seconds). The normalized timing of foot contact and ball release were also calculated and expressed as a percentage of the pitching cycle duration.

Statistical Analysis

Preliminary analyses on ball speed (ie, repeated-measures analysis of variance, with post hoc t tests and Bonferroni correction) revealed that peak pitching performance occurred in the second set rather than during the first set. Therefore, the second set was considered as the pitcher’s peak, or “baseline” performance, while the pitches of the final set were analyzed to represent the pitcher’s late-inning performance at the end of the game-length pitch count. The last 5 pitches of the second and final sets were analyzed. Paired t tests were performed to compare stride length, angular excursions, angular velocities, event timing, and performance outcomes between the baseline and final sets of pitches (SPSS Statistics v22; IBM Corp). Comparisons for the angular excursions were only completed when differences were ≥2° between the baseline and final sets. This cut-off was used to limit the comparisons between sets to those of more clinical significance. Significance was set at P < .05.

Maximum Angular Excursions

For the leading leg, there were no significant differences throughout the stride phase for any maximum angular excursions between the baseline and final sets. For the trailing limb, maximum hip flexion and extension decreased, maximum hip internal rotation increased, and the magnitude of ankle plantar flexion decreased (Table 1).

Angular Excursions at Events

At the balance position, there were no significant differences in the angular excursions at the hip, knee, or ankle of the leading leg between the baseline and final sets. The hip of the trailing leg was more internally rotated, and the knee of the trailing leg was more flexed at the balance position in the final set compared with the baseline set (Table 2). There was no significant difference in the pelvis orientation at the balance position between the baseline and final sets of pitches (baseline set: –126.2° ± 14.5°; final set: –126.4° ± 14.0°; P = .923).

At foot contact, the hip of the leading leg was less flexed and more abducted in the final set compared with the baseline set (Table 2). No other significant differences were observed in the leading leg at foot contact between the baseline and final sets. The pelvis was significantly (P = .017) more rotated in the transverse plane toward the dominant side at foot contact in the final set (–46.2° ± 11.3°) compared with the baseline set (–39.4° ± 8.5°). That is, for a right-handed pitcher, the pelvis was more rotated in the clockwise direction in the final set.

No significant differences in the angular excursions of the leading leg were observed at ball release between the baseline and final sets. The pelvis orientation at ball release was not significantly different between the baseline and final sets (baseline set: 19.3° ± 10.6°; final set: 18.8° ± 11.2°; P = .676).

Maximum Angular Velocities

During the stride phase, the maximum knee extension angular velocity of the trailing leg decreased by 23.2 deg·s⁻¹ in the final set. No other significant differences in maximum angular velocities throughout the stride phase were observed between the baseline and final sets. During the acceleration phase, there was no significant difference in the maximum knee extension angular velocity of the leading leg between sets.

Event Timing

There were no significant differences in the duration of the pitching cycle (baseline set: 1.30 ± 0.16 s; final set: 1.31 ± 0.12 s; P = .858) and the normalized timing of when foot contact (baseline set: 67.2% ± 3.5% pitching cycle; final set: 66.0% ± 3.1% pitching cycle; P = .078) and ball release (baseline set: 77.8% ± 2.7% pitching cycle; final set: 77.8% ± 2.2% pitching cycle; P = .829) occurred between the baseline and final sets.

Performance Outcomes

Ball speed significantly decreased (P < .001) during the final set (28.3 ± 2.5 m·s⁻¹) compared with the baseline set (29.5 ± 2.5 m·s⁻¹). There were no significant differences in the percentages of strike (P = .379), ball 1 (P = .204), and ball 2 (P = .053) pitches between the baseline set (35.5% ± 14.0%, 45.1% ± 14.8%, and 19.3% ± 10.0%, respectively) and final set (32.8% ± 10.4%, 38.8% ± 15.9%, and 28.4% ± 11.5%, respectively).

Balance Position

During the wind-up, pitchers prepare for the stride by rotating the body away from home plate and raising the leading leg to achieve the balance position. ⁴ In the present study, compared with the baseline set, the pitchers in the final set exhibited less rotation of the foot of the trailing leg away from home plate (P = .054) and a more internally rotated hip of the trailing leg (P = .042), while the pelvis orientation remained unchanged. The change at the hip was necessary to maintain the pelvis orientation because of the less rotated foot at the balance position. The knee and hip of the trailing leg were also more flexed (P = .003 and P = .064, respectively) at the balance position, while the hip of the leading leg was less flexed (P = .082) in the final set of pitches. Ball speed has previously been positively correlated to the braking force produced upon foot contact of the leading leg.^13,15 In the present study, decreased hip flexion of the leading leg and a more flexed trailing leg suggest that the leg was not raised as high and may indicate less ability to store gravitational potential energy to be used later. This relationship requires further investigation.

Stride Phase

After the balance position, the pitchers entered the stride phase to drive the body toward home plate. The stride phase involves an initial flexion phase and a subsequent extension phase for the hip, knee, and ankle of the trailing leg. In the final set, hip flexion of the trailing leg decreased compared with the baseline set. Subsequently, pitchers exhibited decreased magnitudes of hip extension and ankle plantar flexion (Table 1) during the stride phase. The altered balance position may have been one of the factors influencing these changes. Specifically, the more flexed hip and knee of the trailing leg at the balance position may have influenced the reduced excursion of sagittal-plane movement of the trailing leg’s hip during the stride phase. There were also decreases in hip abduction and knee extension angular velocities of the trailing leg by the end of the game-length pitching bout. Hip abduction and knee extension would both contribute to driving the body toward home plate. These decreased angular velocities may thus have important implications on the angular momentum produced by the lower extremities. Therefore, the combination of these decreased angular displacements and angular velocities may have resulted in a diminished drive toward home plate, which in consideration of the kinetic chain could be detrimental to the acceleration of the ball.

Compared with the baseline set, stride length was not significantly different in the final set. This observation is in agreement with a previous study comparing the pitching mechanics of high- and low-velocity pitchers. ¹⁴ However, on average, there was an earlier occurrence of foot contact (P = .078) during the late-inning performance compared with the baseline set, resulting in a shortened duration of the stride phase. The earlier occurrence of foot contact may have been influenced by the reduced hip flexion of the leading leg at the balance position, as the leading leg was not raised as high.

At foot contact, the pelvis was less rotated toward home plate in the final set (ie, for a right-handed pitcher, the pelvis experienced less right transverse rotation). This difference in the pelvis orientation may partially be explained by the earlier occurrence of foot contact in the final set, as there was less time to rotate the pelvis toward home plate. Pitchers may have also been slower to rotate their pelvis toward home plate; however, the angular velocity of the pelvis was not assessed in this study. Altered positioning and timing of the rotation of the pelvis have previously been found to affect the kinetics at the shoulder and elbow, ²⁵ which could have a negative effect on the risk of upper extremity injuries. The positioning of the pelvis upon foot contact has also been identified as an important factor influencing ball speed. ²³ As the pelvis remained more rotated away from home plate at foot contact, there may have been less than optimal timing of the pelvic rotation, which could have affected the ball speed in the final set. ²³ Further investigation is required to analyze whether this change in the orientation of the pelvis affected the rest of the pitching motion.

Kinematic differences were also seen in the hip of the leading leg at foot contact, including less flexion and more abduction during the late-inning performance. In the final set, there appears to have been a chain reaction of kinematic changes, starting at the balance position. Specifically, altered positioning of the hip and foot of the trailing leg at the balance position may have influenced the change in the pelvis orientation upon foot contact, which in turn may have contributed to the change in the hip position of the leading leg upon foot contact. The less flexed hip of the leading leg at the balance position and throughout the stride phase may have also influenced the change in the leading leg’s hip position at foot contact. This links back to the concept of the kinetic chain, whereby improper positioning and movements of each body segment can affect subsequent actions throughout the pitching motion. Therefore, changes in lower extremity kinematics during the early phases of the pitching cycle could upset the rest of the pitching motion. Whether these kinematic changes contributed to the decreased ball speed and/or pitching accuracy requires further investigation. If such a relationship is found between these kinematic and performance outcome parameters, the importance of achieving a correct balance position may need to be stressed in adolescent pitchers.

Acceleration Phase

Because of large braking forces experienced upon foot contact, the forward rotation of the shank rapidly decelerates, while inertia maintains the linear movement of the upper body.^11,16 The combined actions of these movements should result in knee extension of the leading leg. Slower ball speeds have previously been associated with increased knee flexion at foot contact and deceased knee extension velocity throughout the acceleration phase.^3,7,14 Although ball speeds decreased in this study, there were no differences in the maximum knee extension angular velocity of the leading leg during the acceleration phase or the magnitude of knee flexion at ball release. Therefore, the kinematic changes in the present study were different to those previously seen in adult pitchers.^6,18 That is, the adolescent pitchers exhibited kinematic changes in the lower extremities primarily at the balance position and during the stride phase rather than at ball release. Whether these kinematic changes that occur earlier in the pitching motion affect trunk and upper extremity movements used to accelerate the ball remains to be determined.

Limitations

A small sample of 12 high school baseball pitchers was enrolled in this pilot study. Although additional participants may have been beneficial, the statistical analyses focused upon intraparticipant differences between the baseline and late-inning performances. While on average the kinematic differences were small, statistical differences were detected for several of the outcome variables. Pitchers may have needed to complete extra warm-up pitches as the peak ball speed was seen in the second set rather than in the first set. Consequently, the late-inning performance was analyzed after 5 sets of pitches following the baseline performance rather than 6 sets. It is possible that by completing an extra set of pitches, larger differences may have been elicited between the baseline and final sets of pitches. However, the results from the present study are still relevant and highlight parameters that are susceptible to change near the end of a game-length pitch count. Further investigation into the effects of these changes on performance and injury risk may be required.

One participant only completed 5 sets because of shoulder pain experienced by the end of the fifth set, so the last 5 pitches of the fifth set were analyzed as their late-inning performance. However, the inclusion of these data did not conflict with the purpose of the study, which was to investigate differences in lower extremity kinematics between baseline and late-inning performances. Kinematic differences between pitchers may have been influenced by their training status and thus their susceptibility to experience fatigue in the lower extremities. However, muscular strength test findings or subjective ratings of perceived effort were not collected. This limits our ability to determine whether the changes in kinematics and performance outcomes were precisely caused by neuromuscular fatigue. However, the results of the present study demonstrate that adolescent pitchers are unable to maintain consistent lower extremity kinematics over a game-length pitch count. Further research is required to confirm whether these kinematic changes in adolescent pitchers are specifically experienced because of muscular fatigue in the lower extremities throughout a game.

Clinical Implications

The combination of changes in lower extremity kinematics, decreased ball speeds, and tendency to throw more wild pitches during the late-inning performance suggests that some form of fatigue may have been experienced by the adolescent pitchers. Compensatory movements further up the kinetic chain (ie, at the pelvis, trunk, and upper extremities) may have been required to contribute to the acceleration of the ball in an effort to maintain ball speed. However, the changes in lower extremity kinematics could prevent proper positioning and timing of the pelvis, trunk, and upper extremity movements, which would be detrimental to the acceleration of the ball. The most notable change by the end of the pitching bout was the orientation of the pelvis at foot contact, which remained more rotated away from home plate. If pitchers begin to exhibit this late rotation of the pelvis during training sessions, coaches should look for noticeable changes in the movements of the lower limbs used to achieve the balance position. In particular, consistent rotation of the foot and hip used to rotate the pelvis away from home plate and the magnitude to which the leading leg is raised may help maintain proper orientation of the pelvis at foot contact. If subsequent research confirms that muscular fatigue is responsible for the kinematic changes, then targeted training strategies (eg, strengthening, power training, maintaining the balance position with extended play) may be employed to improve late-inning lower extremity kinematics and ball speed. ¹⁹ Additional research that correlates lower extremity with trunk and upper extremity function in high school baseball pitching is needed to further understand the role of the lower limbs in fatigue, performance, and injury prevention while pitching.