1 Introduction

Injuries to professional athletes can result in detrimental outcomes with respect to performance, time lost, and quality of life.^1,2 In professional baseball, the number of injuries, number of surgeries performed on injured players, and costs associated with these injuries have all steadily increased over the last several decades.³ Injuries to the core and hip/groin are common, accounting for about one fifth of all players on the Major League Baseball (MLB) injured list (IL).⁴ Pitchers, specifically, experience significantly higher rates of core and groin injuries combined than do position players.⁵ Core and hip/groin injuries can be severely debilitating, causing pitchers to miss about six times as many games as position players.⁶ On average, MLB pitchers with core and hip injuries remain on the IL for five to seven weeks.⁴

Previous work has quantified the impact of prior injury on changes in pitcher kinematics.^7–10 Compared to uninjured pitchers, collegiate and professional pitchers with a history of a superior labrum anterior-posterior (SLAP) tear exhibit significant reductions in shoulder horizontal abduction, shoulder external rotation, and forward trunk tilt.⁷ Collegiate and high school pitchers who have undergone ulnar collateral ligament (UCL) repair exhibit significant reductions in elbow extension, elbow extension velocity, and shoulder internal rotation velocity.¹⁰

The lumbopelvic region in particular, including core musculature, has been of particular interest due to its position and function as a bridge connecting the lower body segments to the throwing arm.¹¹ The core muscles influence pelvic girdle motion in the sagittal, coronal, and transverse planes.¹² Researchers have hypothesized that these muscles provide a stable base for distal motion, providing a platform against which these distal segments can accelerate.¹³ Core injury in MLB pitchers has been linked to a reduced fastball velocity the year of injury,⁴ while poor lumbopelvic control in professional pitchers has been associated with an increased likelihood of missing more than 30 days in a season.¹² Chaudhari et al.¹¹ determined lumbopelvic control in minor league pitchers correlated with fewer walks plus hits per inning pitched in addition to greater innings pitched. In collegiate pitchers and minor league pitchers, poor lumbopelvic control has been found to increase the kinetic forces of shoulder horizontal torque and elbow valgus torque.¹⁴ Even more, hip- and groin-related injuries in MLB pitchers have been found to be more prevalent in pitchers with UCL tears, possibly increasing the likelihood of these injuries.¹⁵ Hence, there is adequate reason to believe core injuries have a clinically significant influence on pitcher kinematics, potentially influencing execution of the throwing motion.

Biomechanical evaluation of pitching kinetics as surrogates for ligament loading has suggested select pitching positions may stress the joint structures of the throwing arm more than others.¹⁶ Trunk motion, and trunk tilt in particular, is intimately related with shoulder kinematics in the throwing arms of pitchers.^17,18 Matuso et al.¹⁷ found that as contralateral trunk tilt increased, the shoulder abduction angle producing the minimum peak elbow varus torque decreased. High school pitchers with improper trunk rotation sequences have demonstrated greater shoulder external rotation and greater shoulder proximal forces.¹⁶ Given the intimacies of throwing arm motion with trunk kinematics, there is a need to further assess how these kinetic and kinematic measures are modulated by preceding core and groin injuries. Therefore, the purpose of this study was to investigate the differences in pelvic, trunk, and throwing arm kinematics in professional pitchers with prior core or groin injuries compared to professional pitchers without prior core injury. The authors hypothesized pitchers with prior core/groin injury would demonstrate increased elbow varus torque and shoulder horizontal adduction torque than pitchers with no prior injury during the throwing motion.

2 Methods

3 Statistical analysis

A Propensity Score Matching Method was utilized to compare the pitching populations and account for any potential confounders.^22,23 Each pitcher in the GCG subgroup was matched in a 4:1 comparison from a pool of pitchers ultimately selected for the NPI subgroup, using the nearest neighbor matching by way of a greedy algorithm.²⁴ Pitchers were matched based on the following confounders: age, height, weight, and handedness. The propensity score for each pitcher was calculated using logistic regression which is defined as follows:logPr(Treatment)Pr(Control)=β0+β1xweight+β2xheight

The generated score is equal to the probability of any pitcher being in the General Core or Groin subgroup based on the aforementioned confounders.

Pitchers in each of the aforementioned injury subgroupings were compared to the NPI group by averaging values among all pitches thrown per pitcher. Descriptive statistics and independent t-tests were then used to compare the subgroupings for the following 20 kinematic variables: Stride length, pelvis rotation, pelvis tilt, pelvis obliquity, trunk flexion, trunk lateral flexion, trunk rotation, trunk tilt, trunk obliquity, lead hip flexion, back hip flexion, lead hip internal rotation, back hip internal rotation, elbow flexion, shoulder abduction, shoulder horizontal adduction, shoulder external rotation, arm slot, lead knee flexion, and ball velocity. During the arm cocking phase, the following peak kinetic values were compared between subgroups: shoulder internal rotation torque, shoulder horizontal adduction torque, shoulder superior force, shoulder anterior force, elbow varus torque, and elbow medial force. Peak elbow anterior force and elbow flexion torque were compared during the arm acceleration phase. Peak shoulder adduction torque, shoulder distractive force, and elbow distractive force were compared during the arm deceleration phase. For all analysis, statistical significance was set at an alpha value of 0.05. All data analysis was performed using R (version 3.6.3, The R Foundation, Vienna, Austria).

4 Results

The incidence of groin and core injuries included in this study is shown in Table 1. The average time from injury to testing for all pitchers was 19.6 ± 22.5 months (range: 12.1 months–46.9 months). Pitchers missed on average, 1.8 ± 1.6 months (range: 0.5–3.0 months) due to their injury and only 1 pitcher required surgical intervention for their injury (inguinal hernia repair).

Professional pitchers were included in the following subgroupings: NPI (n = 76), GCG (n = 19), CM/ST (n = 10), and Spine or Back (n = 8). No significant difference was noted for characteristics utilized in the propensity score calculator (Table 2), with a propensity score of 0.136 calculated for NPI to GCG 4:1 matching.

Kinematic parameters were compared among all injury subgroupings for variables specifically at foot contact and at ball release (Table 3). The NPI subgroup did not significantly differ in pelvic, trunk, or hip kinematics at foot contact compared to the GCG and Spine or Back subgroups. Elbow flexion at ball release did differ between the NPI subgroup and the CM/ST subgroup (31 ± 5 vs. 35 ± 6° respectfully, p = 0.044) while it did not differ for the GCG group (33 ± 5°, p = 0.067) or the Spine or Back subgroup (35 ± 4°, p = 0.063). The CM/ST subgroup had significantly less shoulder abduction than the NPI cohort (88 ± 8 vs. 94 ± 8° respectfully, p = 0.037). Shoulder external rotation at ball release did not differ between groups (NPI: 85 ± 13°; GCG: 84 ± 13°, p = 0.704; CM/ST: 85 ± 10°, p = 0.775; Spine or Back: 87 ± 16°, p = 0.700), nor did ball velocity (NPI: 38.9 ± 1.6 m/s, GCG: 38.3 ± 2.0 m/s, CM/ST: 38.4 ± 2.1 m/s, Spine or Back: 38.8 ± 1.2 m/s; p ≥ 0.316).

Differences in peak kinetic values were noted between injury groups (Table 4), in particular, at arm acceleration. The GCG injury group had significantly greater non-normalized (410.3 ± 53.9 vs. 371.2 ± 57.9 N•m, p = 0.005) and normalized elbow anterior force (43.9 ± 4.7 vs. 40.0 ± 5.2 %BW, p = 0.006) than the NPI subgroup. The GCG subgroup also had significantly greater non-normalized (77.4 ± 11.4 vs. 68.3 ± 11.3 N•m, p = 0.002) and normalized (4.3 ± 0.5 vs. 3.8 ± 0.5 %BWxBH, p = 0.001) elbow flexion torque than the NPI pitchers. The CM/ST subgroup had significantly greater normalized elbow anterior force (44.5 ± 4.3 vs. 40.0 ± 5.2 %BW, p = 0.031), elbow flexion torque (4.4 ± 0.4 vs. 3.8 ± 0.5 %BWxBH, p = 0.002) and normalized (9.5 ± 1.4 vs. 7.9 ± 1.9 %BWxBH, p = 0.007) as well as non-normalized (170.1 ± 34.8 vs. 141.6 ± 38.3 N•m, p = 0.019) shoulder adduction torque than the NPI pitchers. The Spine or Back subgroup had no significant differences in kinetics compared to NPI (p > 0.05).

5 Discussion

Injury is frequently subject of concern for elite sporting teams for a multitude of reasons including effects on performance, time lost, and financial implications. While other groups have evaluated baseball pitchers with prior shoulder and/or elbow injury, the role core injuries play on pitching mechanics as well as upper extremity joint loading remains less examined, and as such, was the purpose of this study. The major findings of this report are as follows: (1) the GCG injury group had significantly greater non-normalized and normalized elbow anterior force and elbow flexion torque than the NPI subgroup; (2) the CM/ST subgroup had significantly greater normalized elbow anterior force, elbow flexion torque, and normalized shoulder adduction torque than the NPI pitchers; and (3) the CM/ST subgroup had significantly greater elbow flexion at ball release compared to the NPI subgroup.

GCG pitchers demonstrated significantly higher peak elbow anterior force and peak elbow flexion torque than pitchers with no prior injury, with no difference in ball velocity outcomes. Less kinetically conservative throws may be explained by the role the legs and trunk serve as the main force generators of the kinetic chain.²⁵ Kibler and Chandler determined that a 20% drop in kinetic energy transferred from the hip and trunk to the arm necessitates a 34% increase in shoulder rotational velocity to transmit the same amount of force to the hand.^26,27 It is certainly possible that pitchers with a prior core/groin injury may be incapable of utilizing the maximum rotational and flexion capacity of the pelvis and trunk as a component of the kinetic chain, thus requiring compensation at the level of the throwing arm. Furthermore, it is also plausible that these improper kinematics existed prior to injury and may have contributed to the injuries in the first place; future studies can address this concern with a prospective trial design assessing pitcher biomechanics before and after injury. A limitation of the current study is the self-reporting bias of player injuries, which did not elucidate if these injuries were incurred while pitching or during some other recreational activity. In addition, whether these kinetic differences were the consequence or cause of injury is unclear and warrants additional investigation.

Increased peak elbow flexion torque has been implicated in pathology of the biceps/labrum, and as such, could play a role in injury risk. Due to the function the biceps brachii serves in producing flexion torque, it has previously been proposed as a risk factor for SLAP tears of the glenoid fossa as well as bicep tendinopathy.²⁸ Through the connection of the long head of the biceps brachii to the anterosuperior glenoid labrum, the muscular contraction necessary to create elbow flexion torque can ultimately create strain on the biceps tendon labrum complex. Ultimately, given the GCG and CM/ST group demonstrated higher elbow flexion torque and shear anterior force at the elbow, these pitchers may be demonstrating less kinetically conservative mechanics that could theoretically place them at further risk of elbow injury. Though these kinetics have not been directly linked with injury, they have been implicated in injury risk at the elbow when used as surrogates for ligament loading.²⁸

GCG pitchers (p = 0.067) as well as the CM/ST subgroup (p = 0.044) achieved higher elbow flexion at ball release compared to NPI pitchers, which may suggest the mechanism by which these pitchers demonstrated elevated peak kinetics at the elbow. Solomito et al.²⁹ previously demonstrated that increased elbow flexion post-ball release correlated with increased elbow varus torque for fastball and curveball pitches in collegiate pitchers, without significant association with ball velocity. Furthermore, cadaveric investigations have shown that when shifting from extension to flexion, loading on the elbow joint is redistributed.³⁰ The UCL receives 31% of total joint force in full elbow extension, while the bony components and other soft tissue components receive the remaining 69%.³¹ The UCL's role to resisting torque climbs to 54% when the elbow is flexed to 90°. As such, it is possible that pitchers with prior groin or core injuries may be more susceptible to injury at the elbow by means of this compensatory mechanism to derive faster throwing arm velocity and ball velocity. In other words, a history of core injury may result in a higher incidence of elbow mechanics that have previously been suggested as predisposing pitchers to injury^32–34 as a means to maintain peak performance. Interestingly, Werner et al.³⁵ noted a positive association between elbow flexion with ball velocity in collegiate pitchers, though this was at foot contact rather than at ball release.

The findings of this study help to build upon concepts described in the current literature. Marshall et al.⁴ noted significantly reduced fastball velocity (91.6 vs 92.2 mph, p = 0.001) for MLB pitchers with core injuries sustained in the same year, while our study noted no difference in ball velocity for professionals with prior injury (Table 3). Though differences in data collection should be noted, this may imply that while performance detriments from core injury may impact a pitcher the year of play, in the long-term, these pitchers are able to return to a previous level of performance. Next, though our study did not directly assess lumbopelvic control, based on prior literature, we hypothesize that a history of core injury may also result in deceased lumbopelvic control. Chaudhari et al.¹² noted for professional pitchers with poor lumbopelvic control a significantly higher number of days missed due to injury. Laudner et al.¹⁴ also found increased shoulder horizontal torque and elbow valgus torque in collegiate and minor league pitchers with poor lumbopelvic control, similar to our studies findings of elevated peak kinetics for pitchers with prior core injury. Most interestingly, Kantrowitz et al.¹⁵ found MLB pitchers with hip- and groin-related injuries have a higher prevalence of UCL tears, suggesting core injuries may potentially increase the likelihood of these injuries. These reports together in the context of our study support the notion that core/groin injuries as well as poor lumbopelvic control may increase compensation at the level of the throwing arm, subsequently placing a pitcher at higher risk of elbow injury. Ultimately, additional investigation examining the role of core injury and instability in the context of throwing arm injury and kinetic loading is warranted.

One parameter not investigated in the current study, but may be a contributor to the theorized increasing throwing arm compensation in pitchers with prior core injury, is limitations in the passive range of motion (ROM) of the hip. Li et al.³⁶ observed in a cohort of 29 professional baseball pitchers with hip, groin, and hamstring injury decreased hip rotation arc when compared to a non-injured cohort. The functional hip ROM required by the leading leg of the professional baseball pitcher has been noted to exceed that of the stance leg, suggesting an importance of this physical examination in delineating optimal pitching biomechanics specific to the throwing side.³⁷ Limitations in lead leg hip internal rotation could reduce the lower extremity's capacity to absorb force, requiring the posterior rotator cuff to eccentrically absorb loads and delay the arm at ball release.^28,38–41 These compensations may increase shear forces at the anterior glenohumeral joint and medial elbow. Future biomechanical analysis should consider incorporating these physical examination findings as potential contributors to differed throwing mechanics.

This is the first study to the authors' knowledge evaluating pitchers with prior injury of the core, groin, and spine/back for potential differences in kinetic and kinematic parameters during the pitching motion. The strengths of this study are that the injury data is standardized by region, timing, and severity (utilizing time out of play as a surrogate for severity of injury), along with a 4:1 propensity match scoring to adequately control for confounding variables such as height, weight, age, and handedness. Additionally, this study utilized a higher frame capture rate than prior motion-capture study evaluations (480 Hz vs 240 Hz). Notably, this study's method of injury-history collection, compared to prior database studies within Major League Baseball,^42–46 has the distinct advantage of spanning 10 years rather than solely a pitchers tenure as a professional. Limitations of this study were also present. An a-priori exam determined a minimum 21 injured pitchers would be sufficient for a well powered analysis. Therefore, our GCG group was slightly underpowered (n = 19). The CM/ST as well as the Spine or Back subgroups had an even smaller n potentially presenting type II error; as such, there may in fact be an association between these injuries and detrimental kinematics when compared to their uninjured peers. As mentioned previously, pitchers' injuries were self-reported and intrinsically subject to reporter bias. Finally, this study only evaluated professional baseball pitchers in the United States MLB; therefore, its applicability to other pitching cohorts (ie. lower playing levels, youth/adolescent pitchers, other playing positions, non-MLB professional league pitchers, etc.) is unclear.

6 Conclusion

Professional baseball pitchers with prior core/groin injuries demonstrated increased elbow anterior force and elbow flexion torque compared to pitchers with no prior injury. It is possible these pitchers inadequately utilize their lower extremities as a component of the kinetic chain, subsequently compensating at the level of the throwing arm. This in turn, has the theoretical potential to place them at risk of elbow injury. Whether these kinetic differences result as a consequence of injury or serve as a risk of such warrants additional investigation.