Strength and conditioning professionals routinely discuss the importance trunk training and its effect on performance among many different sports. More specifically, the importance of trunk strength and endurance in overarm throwing is suggested to be vital for both performance outcomes and injury prevention (4,5,8,9). These studies have also indicated that trunk contributions are vital for both the demonstration of high throwing velocity and for improving throwing velocity within individual pitchers. Thus, the notion of enhancing trunk performance characteristics in throwing, via strength training, may augment the process of attempting to improve throwing velocity. However, no research to date has examined the efficacy of trunk training and its effects on throwing performance.
Elite throwers generate high angular velocities at both the pelvis (662 ± 148°/s) and upper torso (1180 ± 294°/s), and exhibit approximately 50-60° of differentiated rotation between the pelvis and upper torso (2,4,9). In general, more advanced pitchers achieve higher pelvis and upper torso velocities than less-developed pitchers (6). Additionally, increases in ball velocity within pitchers are associated with increased pelvis and upper torso velocities (9). Theoretically, increased pelvis and upper torso velocities would allow more energy to be transferred from the trunk to the throwing arm, and eventually to the ball, which will lead to an increase in ball velocity (9). Specific training that focuses on improving both range of motion and velocities of the trunk would seem to be important for augmenting throwing velocity. Furthermore, integrating specific trunk exercises that demonstrate similar ranges of motion and velocities produced in throwing may ultimately lead to higher throwing velocities. However, no research has addressed the idea of attempting to alter trunk kinematics via specific trunk training and determine its effect on trunk performance characteristics and ball velocity. Additionally, no study has examined trunk kinematics in typical trunk training exercises that are implemented for training purposes to examine their assumed efficacy for promoting a training effect. This directs us to the question: Do current trunk training exercises demonstrate appropriate pelvis and upper torso range of motion and velocities that would be specific enough to promote improvements in performance characteristics?
The purpose of this study was to examine the maximum differential trunk rotation and maximum angular velocities of the pelvis and upper torso of athletes while they performed 4 trunk exercises (seated band rotations, cross-overs, medicine ball throws, and twisters) and compare these trunk exercise kinematics with the trunk kinematics demonstrated in actual throwing performance.
Trunk musculature important for throwing performance include the rectus abdominis, external oblique, internal oblique, and transverses abdominis. The roles of this musculature in throwing are to promote dynamic stabilization, rotation, lateral bending, and flexion/extension of the trunk. Improper training or a lack of training could lead to muscle imbalances and injuries, not only to the trunk, but also to the upper extremities (10). In addition, appropriate training of the trunk musculature may help to improve sport performance. To optimize the contribution of the trunk in throwing, athletes must be able to effectively use energy generated by the trunk musculature and optimally transfer that energy through their system (1). Maintaining a strong trunk may also decrease the force demanded by the musculature associated with the shoulder and elbow joints to produce velocity (8,9). Applications of kinetic chain theory, neuromuscular contributions from stretch-shortening cycles, as well as proximal to distal sequencing all support the importance of training in a full range of motion for optimal performance (3,5).
Proper training of trunk musculature should focus on increasing range of motion, strength, endurance, and velocity potential (5). Training for maximum differentiated trunk rotation and maximum angular pelvis and upper torso velocity will help to develop increased force through a greater range of motion (7). Increased forces generated by trunk musculature will likely produce higher trunk velocities and, more specifically, pitching velocities. This can be accomplished through sport specific training.
Experimental Approach to the Problem
Pelvis and upper torso kinematics, including the amount of differentiated rotation between the pelvis and upper torso and maximum angular velocities, were examined in 4 specific trunk training exercises and compared with pelvis and upper torso kinematics in throwing performance. Kinematics of trunk training exercises, to our knowledge, have not been previously examined. These data may provide an estimation of the efficacy of trunk training exercises with respect to the possibility of producing a specific training effect for improving throwing velocity.
Nine NCAA Division I male baseball players (8 pitchers and 1 position player) participated in this study. The participants average age was 19.7 ± 1.0 yrs, height of 187.0 ± 5.1 cm, and a mass of 84.7 ± 9.7 kg. Informed consent was obtained before participation was allowed. All participants were healthy and had no training restrictions. Six of the participants were right hand dominant, and 3 were left hand dominant. Dominance was defined by the individuals throwing arm preference. All participants had at least 1 year of training experience with the exercises, as they were included in their regular strength and conditioning training program at the University.
Testing was conducted during the middle of the participants' competitive season. Participants wore spandex shorts and no shirts to aid in marking the correct anatomical landmarks. Retroreflective markers were placed bilaterally on the lateral tip of the acromion process and the anterior superior iliac spine of each participant. After completing the informed consent and health history forms, participants were taken through a general and specific warm-up. The specific warm-up included performing each of the trunk exercises to be tested.
The pelvis vector was defined by a line connecting the 2 pelvis markers. The upper torso vector was defined by a line connecting the shoulder markers. Maximum differentiated rotation was defined by the maximum angle difference between the pelvis vector and upper torso vector in the XY plane (transverse plane). Kinematic data were collected using a 4-camera 3-dimensional video system (Motion Analysis Corporation, Santa Rosa, Calif.). Video data was sampled at 60 Hz. (Eva 7.0 software, Motion Analysis Corporation. Santa Rosa, Calif.). Maximum angular velocities of the pelvis and upper torso in the XY plane were calculated using Kintrac software (Motion Analysis Corporation, Santa Rosa, Calif.). Angular velocity of the pelvis was the cross product of the pelvis vector and its derivative. Angular velocity of the upper torso was the cross product of the upper torso vector and its derivative (Kintrac, Motion Analysis Corporation, Santa Rosa, Calif.).
The 4 trunk exercises tested in this study represent different types of trunk training exercises for promoting range of motion, rotational velocity, and explosiveness. The exercises were as follows: (i) seated band rotations, (ii) lying cross-over crunches, (iii) standing medicine ball throws from the hip, and (iv) twisters. The purpose of the seated band trunk rotations and the lying cross-over crunches was to demonstrate maximum range of motion within the limitations of the exercise and thus, demonstrate maximum differentiated rotation of the pelvis and upper torso. The purpose of the medicine ball throw was to produce the greatest horizontal distance thrown. The purpose of the twister was to have the subject rotate the trunk as quickly as possible while still placing an emphasis on rotating the upper torso as much as possible. Maximum differentiated trunk rotation was examined in all four exercises. Maximum upper torso velocities were calculated only for twisters and the medicine ball throw because they were the only 2 exercises where velocity was critical for performing the exercise. Maximum pelvis velocities were examined only in the medicine ball throws because of the fact that the pelvis was primarily fixed to the ground in the twister exercise. Participants performed the exercises in random order.
Seated band rotations required participants to hold the end of an exercise band that was tied around a fixed object and rotate the trunk (Figure 1). The participant's elbows remained fully extended in front of his body and parallel to the ground throughout the exercise. From this initial position, participants rotated the upper torso as far as possible away from the fixed object, thus increasing the tension of the band as it lengthened. Participants proceeded to rotate the trunk back into the starting position and performed the movement again until 5 repetitions were completed. Individuals also performed the exercise 5 times in the opposite direction.
Lying cross-over exercise required participants to lie in a supine position on the ground (Figure 2). To perform right directional rotation, the subject's left foot was flat on the ground with the left knee bent approximately 90° and the right ankle was placed over the flexed left knee. The right arm was flat on the ground beside the right hip and the left hand was placed behind the head. From this position, participants performed the exercise by touching the left elbow to the right knee and then returned to the starting position while the left hand remained behind the head. Individuals were instructed to rotate as far as possible keeping the right knee stationary. This exercise was completed 5 times in one direction. Participants then completed the exercise 5 times in the opposite direction.
Standing medicine ball throws required participants to hold a medicine ball at mid-chest level (Figure 3). From this position, participants were instructed to rotate the body away from the directed line of the throw, producing a countermovement, then rotate in the opposite direction and throw the medicine ball for maximum horizontal distance. Additional instructions were to keep the elbows extended and maintain shoulder flexion at approximately mid-chest height. Participants were told to throw the medicine ball with maximum effort. Medicine balls ranged in mass from 3.7-4.5 kg. This exercise was performed 5 times in each direction.
The twister required participants to sit on the ground with the torso at approximately 60° to the floor with the feet approximately 10-15 cm off the ground and the knees bent approximately 90° (Figure 4). Participants held a 4.5-kg plate on their chest with both hands. Participants were then told to rotate their upper body as far and as fast as possible from side to side for 5 complete rotations.
All data were averaged and categorized into dominant and nondominant rotational directions based on each participant's dominant throwing arm. Average maximum differentiated rotation for band rotations, cross-overs, and twisters were presented as dominant and nondominant directions. Dominant rotation in the exercises for left-handed throwers was band trunk rotations to the right, cross-over to the right, and twisters to the right. Dominant rotation in the exercises for right-handed throwers was band trunk rotations to the left, crossovers to the left, and twisters to the left. Differentiated rotation for the medicine ball throws are presented as preparatory (countermovement) and throwing movement directions. For right-handed throwers, the preparatory movement was right rotation and the throwing direction was left rotation. For left-handed throwers, the preparatory movement was left rotation and the throwing direction was right rotation. Maximum pelvis and upper torso angular velocities for the medicine ball throws were calculated only in the throwing direction. Maximum upper torso angular velocities for the twister were calculated for both rotational movement directions.
Descriptive statistics for average maximum pelvis and upper torso differentiated rotation and average maximum pelvis and upper torso angular velocities are presented in Figures 5 and 6. Intraclass correlation coefficients were calculated to determine the internal consistency of all measures. The calculated values for all measures were between 0.8963 and 0.9913.
Of the 4 exercises tested, seated band rotations produced the greatest amount of differential trunk rotation (dominant = 61.6 ± 6.8°, nondominant = 59.4 ± 7.5°) and cross-overs elicited the least amount of trunk rotation (dominant = 45.2 ± 7.1°, nondominant = 47.0 ± 9.2°). Medicine ball throws demonstrated 50 ± 7.6° (dominant prep), 55.2 ± 9.2° (nondominant prep), and 50.2 ± 7.2° (dominant follow through), 43.9 ± 8.9° (nondominant follow through) of maximum differential trunk rotation, whereas the twisters demonstrated 52.8 ± 12.3° (dominant) and 54.3 ± 16.4° (nondominant) of maximum differential trunk rotation.
A second purpose of the twisters and medicine ball throws was to obtain maximal trunk angular velocity (Figure 6). Average maximum upper torso angular velocities for the twisters were 436.7 ± 101.9°/s (dominant), 415.5 ± 85.5°/s (nondominant). The average maximum angular velocity of the upper torso while performing the medicine ball throw was 493.7 ± 45.0°/s (dominant) and 450.7 ± 61.6°/s (nondominant). The average maximum angular velocity of the pelvis while performing the medicine ball throw was 293.6 ± 73.5°/s (dominant) and 301.6 ± 64.6°/s (nondominant).
The purpose of this study was to examine the maximum differential trunk rotation and maximum angular velocities of the pelvis and upper torso in athletes while they performed 4 trunk exercises (seated band rotations, cross-overs, medicine ball throws, and twisters) and compare these trunk exercise kinematics with the trunk kinematics that are demonstrated in actual throwing performance.
Previous studies have indicated that pitchers demonstrate maximum differentiated trunk rotation of approximately 47-60° (4,9). Results of this study indicated all 4 trunk exercises demonstrated differential rotation that approximated the differentiated rotation in throwing performance. The cross-overs produced the least amount of differentiated trunk rotation of all the exercises tested. This was probably caused by the anatomical restrictions when performing a cross-over. The anatomical structure of the vertebrae most likely limited the amount of rotation within the vertebral column when the spine was flexed.
The goal of the twister exercise was to obtain maximum upper torso velocity and the goal of the medicine ball throw was to produce maximum horizontal distance; thus, the focus of these 2 exercises was not necessarily on maximizing differentiated rotation. However, both twisters and medicine ball throws achieved greater differential trunk rotation than the cross-overs. Therefore, cross-overs were determined to be the least effective of all 4 exercises with respect to maximum differentiated rotation. However, because trunk motion in throwing is multiplanar, the crossover exercise exhibits motion that corresponds to similar trunk movement demonstrated in throwing. Overall, all of the exercises tested produced range of motion that was similar to the range of motion produced in throwing performance and seem sufficient to promote the range of motion necessary for throwing performance capabilities.
Research conducted with elite baseball pitchers has demonstrated upper torso angular velocities of approximately 1100 to 1300°/s during the pitching motion (2,6,9). These same studies have also indicated that the pelvis achieves maximum angular velocities of approximately 500 to 700°/s. Results of this study show that maximum upper torso angular velocities in the exercises where the primary goal was trunk velocity or explosiveness were less than 50% of maximum upper torso angular velocities exhibited in throwing performance. Maximum pelvis angular velocities in the medicine ball throws were approximately 50% as compared with what is demonstrated in throwing performance.
Two aspects of the trunk training exercises that most likely affected the trunk angular velocities in the explosive medicine ball throws were the masses of the medicine balls used, which were between 3.7 kg and 4.5 kg (8-10 lbs) and the nature of the movement where the arm are extended in front of the body at approximately mid-chest height. The medicine ball masses were chosen because they are typical masses that are used in explosive type training. This mass, along with the nature of the movement where the mass is approximately 36 to 46 cm away from the trunk, would most likely lead to an increase in muscle stiffness within the trunk and upper extremities throughout the entire movement. This would occur to both maintain the height of the medicine ball throughout the entire range of motion (ROM) of the exercise and to compensate for the increased inertial forces associated with the mass of the medicine ball, distance from the trunk axis of rotation, and acceleration and deceleration aspects of the countermovement and acceleration phases of the exercise. In throwing, the body compensates for (i.e., temporarily reduces) these inertial characteristics by flexing the elbow during the windup phase of throwing, which reduces the moment of inertia about the longitudinal axis of the trunk. The body also reduces the inertial effect of the upper extremity during trunk rotation through the passive external rotation at the glenohumeral joint during the arm cocking phase of throwing, which is where the pelvis and upper torso reach maximum angular velocities.
The effect of increased mass and size of balls thrown on pelvis and upper torso angular velocities between football throwing and baseball pitching suggests the same pattern of decreasing angular velocities as demonstrated in these results (2). Fleisig et al. (2) showed that pelvis angular velocities decreased approximately 24% (660°/s vs. 500°/s) and upper torso angular velocities decreased approximately 19% (1170°/s vs. 950°/s) when comparing baseball pitching and football passing. Although the medicine ball throw in this study is not similar to the actual throwing motion, it provides an estimate of the effects of adding mass to the distal aspects of the kinetic chain system and increasing the distance from the axis of rotation. The twisters represented a somewhat different type of exercise than the ballistic medicine ball throw, but the emphasis on rotational velocity remained the goal. The increased mass that individuals held at their chest probably affected the velocity of the movement simply because of the increased inertial properties, but the repetitive movement of rapidly accelerating and decelerating the trunk may have resulted in a more controlled movement, as compared with the ballistic throw. Thus, the ability to generate maximal velocity with successive directional changes was constrained.
When assessing the efficacy of a training program for athletes where throwing or striking is important, understanding the purpose of the exercises implemented should be addressed. Is the purpose of the trunk training exercises to maintain or promote increased trunk strength, increased ROM, increased stability, increased endurance, or increased explosiveness? The results of this study provide some compelling results as to what the exercises may provide in terms of a training effect. If the purpose of these exercises and many exercises similar to the ones tested in this study were to improve ROM, endurance, or stability, then these exercises would seem to be appropriate. However, if the purpose of the exercises is to promote enhanced trunk rotational velocity or power that would be specific to throwing or striking performance, perhaps these exercises (specifically, the twisters and medicine ball throws) are not as appropriate as they need to be. Utilizing smaller masses of medicine balls or external objects may provide the necessary reduction in inertial characteristics, thus promoting the capability for enhanced rotational velocity during these types of exercises.
In summary, the trunk training exercises tested in this study demonstrated differentiated trunk rotation similar to what is demonstrated in throwing performance. However, maximum pelvis and upper torso angular velocities exhibited in the trunk exercises were generally 50% or less than what is demonstrated during throwing performance. These results suggest that more careful consideration, relating to the specific goal of exercises, should be given when developing a training program for sports that specifically include throwing and striking (e.g., baseball, softball, and tennis). Incorporating trunk exercises into training programs that promote greater differentiated trunk rotation and greater trunk angular velocities that are similar to the sport movement may help to promote increases in throwing velocity. Specifically, exercises that promote explosive trunk rotation, including torso twists with a partner (i.e., Russian twists) or explosive medicine ball explosive throws (as portrayed in the figures), would seem to be the most appropriate exercises to specifically enhance trunk rotational velocities. The focus of these ballistic throw type exercises should be directed towards rotational velocity; thus, a possible way to address this specific focus may be to decrease the mass of the medicine balls used.
On the basis of the results of the limited number and types of exercises included in this study, it would seem that most trunk training exercises generally used in strength and conditioning programs would demonstrate sufficient differentiated trunk rotation, but an emphasis in these exercises should include the production of maximum range of motion. In addition, generating more energy in the trunk may decrease the risk of injury by decreasing the relative contribution of shoulder and elbow musculature that would demanded to compensate for the lack of optimal lower extremity involvement. Further assessment of trunk training exercises and the effects of training on performance is warranted.
1. Dillman, CJ, Fleisig, GS, and Andrews, JR. Biomechanics of pitching with emphasis upon shoulder kinematics. J Orthop Sports Phys Ther
18: 402-408, 1993.
2. Fleisig, GS, Escamilla, RF, Andrews, JR, Matsuo, T, Satterwhite, Y, and Barrentine, SW. Kinematic and kinetic comparison between baseball pitching and football passing. J Appl Biomech
12: 207-224, 1996.
3. Herring, RM and Chapman, AE. Effects of changes in segmental values and timing of both torque and torque reversal in simulated throws. J Biomech
25: 1173-1184, 1992.
4. Hong, D and Roberts, EM. Angular movement characteristics of the upper trunk and hips in skilled baseball pitching. In: Biomechanics in Sports XI: Proceedings of the XI Symposium of the International Society of Biomechanics in Sports
5. Ishida, K and Hirano, Y. Effects of non-throwing arm on trunk and throwing arm movements in baseball pitching. Int. J Sport Health Sci
2: 119-128, 2004.
6. Matsuo, T, Escamilla, RF, Fleisig, GS, Barrentine, SW, and Andrews, JR. Comparison of kinematic and temporal parameters between different pitch velocity groups. J Appl Biomech
17: 1-13, 2001.
7. Neal, RJ, Snyder, CW, and Kroonenberg, PM. Individual differences and segment interactions in throwing. Human Movement Sci
10: 653-679, 1991.
8. Stodden, DF, Fleisig, GS, McLean SP, and Andrews, JR. Relationship of biomechanical factors to baseball pitching velocity: Within pitcher variation. J Appl Biomech
21: 44-56, 2005.
9. Stodden, DF, Fleisig, GS, McLean, SP, Lyman, SL, and Andrews, JR. Relationship of pelvis and upper torso kinematics to pitched baseball velocity. J Appl Biomech
17: 164-172, 2001.
10. Whiteside, JA, Andrews, JR, and Fleisig, GS. Elbow injuries in young baseball players. Phys Sportsmed
27: 87-102, 1999.