Context

How different pitching roles affect the risk of arm injury in professional pitchers is currently unclear.

Objectives

To investigate differences between professional baseball starting and relief pitchers in the hazard of (1) arm injury and (2) elbow and shoulder injury.

Design

Prospective cohort study.

Setting

Minor League Baseball (MiLB) from 2013 to 2019.

Patients or Other Participants

Pitchers in MiLB.

Main Outcome Measure(s)

Pitchers were followed for the entire MiLB season, and athlete-exposures and injuries were recorded. Risk ratios and risk difference were calculated between starting and relieving MiLB pitchers. A Cox survival analysis was then performed in relation to time to arm injury between starting and relieving MiLB pitchers. Subgroup analyses were conducted for elbow and shoulder injuries.

Results

A total of 297 pitchers were included, with 85 270 player-days recorded. The incidence of arm injuries was 11.4 per 10 000 athlete-exposures. Starting pitchers demonstrated a greater risk ratio (1.2 [95% CI = 1.1, 1.3]), risk difference (13.6 [95% CI = 5.6, 21.6]), and hazard of arm injury (2.4 [95% CI = 1.5, 4.0]) than relief pitchers. No differences were observed for the hazard of elbow injury between starting and relief pitchers (1.9; 95% CI = 0.8, 4.2). Starting pitchers had a greater hazard of shoulder injury than relief pitchers (3.8 [95% CI = 2.0, 7.1]).

Conclusions

Starting pitchers displayed a 2.4 times greater hazard of arm injury than relief pitchers. Subgroup analyses indicated that starters exhibited a greater hazard of shoulder injury than relievers, but no differences occurred for the hazard of elbow injury. However, due to the wide CIs, these subgroup analyses should be interpreted with caution. Clinicians may need to consider cumulative exposure and fatigue and how these factors relate to different pitching roles when assessing the risk of pitching arm injury.

Key Points
  • Starting pitchers demonstrated a 2.4 times greater hazard of arm injury than relief pitchers.

  • Subgroup analyses showed that starters exhibited a greater hazard of shoulder injury than relievers, but no differences were observed for time to elbow injury. However, due to the wide CIs, these subgroup analyses should be interpreted with caution.

  • Clinicians may need to consider cumulative exposure and fatigue and how these factors relate to different pitching roles when assessing the risk of pitching arm injury.

Arm injuries in baseball are a substantial problem that has continued to increase over the past decade.1,2  The highest injury incidence in pitchers is to the elbow and shoulder,1,3  with the prevalence of professional baseball ulnar collateral ligament reconstruction reported as high as 16%.3  Due to the substantial health care required4  for baseball pitching injuries and the resulting effects on work,2  specific injury risk factors have been identified, including shoulder range of motion (ROM),5,6  humeral torsion (HT),7  pitch load,8  and body composition.9  However, given the continued rise in pitching injuries,1,2  these recognized injury risk factors are incomplete, signifying the need for further exploration of other possible risk factors.

One factor that could be related to the risk of arm injury is designated pitching roles. Pitchers can be assigned to multiple roles, including starters and relievers. Starters and relievers have different numbers of rest days between appearances, throw different numbers of pitches per outing and throughout the season, and may pitch in different game situations.10  For example, starting pitchers are expected to pitch at least 5 innings at the beginning of the game, whereas relievers usually pitch 1 to 3 innings at the end of the game and perhaps in high-pressure situations. These different pitching roles create possible discrepancies in pitch load,11  shoulder ROM,12  and performance,10  potentially altering the arm injury risk. As a result, we need to understand how different pitching roles relate to the risk of arm injury.

It is currently unclear how different pitching roles affect the arm injury risk in professional pitchers. Understanding this relationship will allow greater precision in physical examination and shared decision making with coaches and athletes when evaluating the injury risk and creating personalized injury-prevention programs. Therefore, the purpose of our study was to investigate differences between professional baseball starting and relief pitchers in the hazards of (1) arm injury and (2) elbow and shoulder injury.

Study Design

We conducted a prospective cohort study from 2013 to 2019 involving Minor League Baseball (MiLB) pitchers in 1 Major League Baseball (MLB) organization. Before data collection, the risks and benefits of study participation were explained to all pitchers, and they gave verbal and written consent. Professional pitchers were tested for shoulder ROM and body mass index (BMI) in the preseason (ie, spring training), before participating in baseball-related activities. They underwent routine physical examinations and injury screening during preseason medical physicals and were then followed for the entire MiLB season. Athlete-exposures (AEs) and injuries were recorded. This study was approved by the Health System Institutional Review Board.

Study Participants

The MiLB pitchers in short season A to AAA were included in this study. Inclusion criteria consisted of pitchers who participated (1) in all baseball-related training, practices, and competitions and (2) only during MiLB spring training. Exclusion criteria consisted of pitchers who (1) participated in MLB spring training (2) signed a professional contract in the middle of the season.

Outcomes

Exposure definition

An AE was defined as 1 athlete participating in 1 practice or competition in which a player was at risk of sustaining an injury.13  Baseball exposure was defined from the beginning of the preseason (ie, spring training) to the end of the MiLB season.14  An athletic trainer with each affiliate collected all exposure and injury data.

Injury Definition

An injury was defined as damage to a tendon, ligament, nerve, muscle, or bone that occurred during any baseball team-sponsored activity or event15  and resulted in at least 1 day of missed practice or game and for which the athlete received medical attention from an athletic trainer or physician.5  If a player was unavailable to play for injury-prevention reasons (ie, had reached league or individually determined limits for the pitch or inning count), then the absence was not considered an injury.14  Injuries were defined by the Orchard Sports Injury Classification system, and arm injuries were stratified by shoulder or clavicle, upper arm, elbow, and forearm.16  All other injuries and illnesses were also recorded and time loss was considered for overall exposure, but these conditions were not included in the injury analyses.17  Injury severity was calculated as the overall time loss from participation in practices or games,17  with further injury severity stratified by 7 (moderate) and 28 (severe) days.

Explanatory Variable

Pitching Position

Our primary explanatory variable of interest was pitching position. Pitchers were dichotomized as either starters or relievers. Because some pitchers may start and relieve in the same season, we hand counted pitching appearances in case of discrepancies. If a pitcher pitched more than 10% of total appearances as a starter or reliever, a consensus was reached through discussion on the proper pitching position for the individual and confirmed by the organization. For example, if 80% of a pitcher's appearances were as a starter and the other 20% were as a long reliever, our discussion with the team coaches confirmed that the pitcher would be deemed a starter for the analyses.

Confounders

In this study, a confounding variable was any participant characteristic or clinical factor that might have distorted the primary relationship of interest between pitching position (starter versus reliever) and arm injury. Not controlling for confounding variables would result in a biased analysis of the relationship between the outcome (ie, arm injury) and the explanatory variable (ie, pitching position). Confounders were identified through clinical reasoning among the study team and a detailed review of the relevant literature. Confounders controlled for were age, BMI,9  arm dominance,18  seasonal pitch load (ie, total number of pitches),8  number of pitching appearances,10,19  dominant-shoulder total ROM (TROM),6  dominant-shoulder horizontal adduction (HA) ROM,5,20,21  dominant-arm versus nondominant-arm HT difference,7  previous arm injury,7  and year of data collection.

Seasonal Pitch Load and Number of Pitching Appearances

The number of pitches thrown and pitching appearances in a season were collected from team reports. However, due to our research agreement with the MLB organization, seasonal pitch counts and pitching appearances were not included in the data until 2016. As a result, we accessed and recorded the prior pitcher season pitch counts and pitching appearances via publicly available data.22,23  Although these publicly available websites recorded the number of pitching appearances, they did not record the actual number of pitches in a season. As a result, indirect calculations were necessary. Box score statistics, including batters faced, strike outs, and walks, were obtained from the publicly available websites and then calculated via a pitch count estimator formula (3.3 × batters faced + 1.5 × strike outs + 2.2 × walks).24  This pitch count formula has been found to have excellent reliability (intraclass correlation coefficient = 0.98), with an error of 3 pitches in a season.24 

Shoulder ROM and HT

All data administrators were blinded to hand dominance.5  Measurements were taken only at the beginning of spring training. Shoulder ROM (external rotation [ER], internal rotation [IR], and HA) was measured in supine position by 2 examiners for both the dominant and nondominant arms using a digital inclinometer per previously described methods.5,20,21  Shoulder ER and IR were summed to measure shoulder TROM. The difference between the dominant and nondominant shoulder ROM was calculated for ER, IR, HA, and TROM.6  The HT for each arm was measured with the participant in supine position, and indirect ultrasonography was performed and a digital inclinometer was used as documented earlier.7,20  Shoulder ROM and HT measures demonstrated excellent intrarater reliability (intraclass correlation coefficient [2,1] = 0.92–0.99) throughout the study.20 

Injury History

Any history of arm injury was obtained from player medical records and previous professional baseball seasons. A pitcher was defined as having a history of arm injury if he had undergone elbow or shoulder surgery or lost ≥15 days because of an elbow or shoulder time-loss injury in previous professional baseball seasons (or both).6 

Statistical Analyses

All data were investigated for missingness before analyses. Missing data were few (shoulder ROM = 3%, age = <1%, position = 0%) and, thus, complete case analyses were performed. Participant statistics were described using mean ± SD for continuous normally distributed variables, median and interquartile range for continuous non-normally distributed variables, and frequencies and percentages for categorical variables. Arm, elbow, and shoulder injury incidences (ie, new injuries) were calculated by summing arm, elbow, or shoulder injuries and dividing by the sum of person-days multiplied by 10 000. Person-days in this context was the calculation of AE. Risk ratios (RRs) and risk difference (RDs) with 95% CIs were determined. The RR provides a relative risk calculation between the exposure group (ie, starting pitchers) and the nonexposure group (ie, relief pitchers). The RD, or attributable risk, provides an absolute risk calculation that is the difference between the exposure group (ie, starting pitchers) and the nonexposed group (ie, relief pitchers).

Nonlinearity

To describe the relationship between pitching position and arm injury risk, we used a regression model building approach, which allowed us to control for the key confounding variables outlined earlier. Before model development, continuous variables were assessed for nonlinearity in relation to time to pitching arm injury (the outcome of interest). Nonlinearity was assessed through restricted cubic splines at 3, 4, and 5 knots.25  A restricted cubic spline is a nonlinear piecewise polynomial (nonlinear calculation) joined at specific knots throughout the data. Knots are quantile mark points in which each segment (between knots) is evaluated for potential nonlinear relationships.25  The range of data is joined at each successive knot, allowing for different nonlinear relationships to be assessed throughout the data.25  We determined that a relationship between the outcome (ie, arm injury) and a continuous confounder was nonlinear based on the Akaike Information Criteria, residuals, visual inspection, and biological plausibility. The BMI demonstrated a linear relationship, whereas seasonal pitch load, number of pitching appearances, and HT difference demonstrated a nonlinear relationship with 4 knots. Dominant TROM and HA also displayed a nonlinear relationship with 3 knots (Appendix). These nonlinear transformations were included in the models.

Primary and Sensitivity Analyses

A Cox proportional hazards survival analysis was performed to determine the relationship between time to arm injury and pitching position (starter versus reliever) in MiLB pitchers. A hazard ratio denotes the instantaneous rate of injury at a given time point. Hazard RRs with 95% CIs were determined. With Cox survival models, we controlled for age, BMI, seasonal pitch load, number of pitching appearances, arm dominance, dominant-shoulder TROM, dominant-shoulder HA, difference between dominant- and nondominant-shoulder HT, previous arm injury, and year of data collection. We conducted subgroup analyses to determine the risk of elbow and shoulder injuries separately based on pitching position. Sensitivity analyses allowed us to understand the stability of the results. The sensitivity analyses involved a Cox proportional hazards survival analysis for which a linearity of confounders was assumed, a Cox proportional hazards survival analysis with arm injury defined as ≥7 days lost to injury (7+), a Cox proportional hazards survival analysis with arm injury defined as ≥28 days lost to injury (28+), and a logistic analysis controlling for confounders implementing both nonlinear and linear analyses. All analyses were conducted in R (version 4.02; R Core Team) using the dplyr package for cleaning and coding, rms package for nonlinear analyses, survival package for survival analyses, epiR package for risk calculations, and survminer and ggplot2 packages for data visualization.

A total of 297 pitchers participated, and 85 270 player-days (overall AE) were recorded. Starting pitchers threw a median of 53.9 (interquartile range = 24.1–83.8) and relief pitchers threw a median of 23.0 (interquartile range = 16.5–29.5) pitches per appearance. The overall (starters and relievers combined) arm injury incidence was 11.4 arm injuries per 10 000 AEs, 4.8 elbow injuries per 10 000 AEs, and 7.9 shoulder injuries per 10 000 AEs (Table 1). Starting pitchers demonstrated an increased RR (1.2 [95% CI = 1.1, 1.3]; P < .001) and RD (13.6; 95% CI = 5.6, 21.6; P < .001) compared with relief pitchers.

Table 1

Pitcher Demographics

Pitcher Demographics
Pitcher Demographics

Hazard of Arm Injury Between Starting and Relief Pitchers

Starting pitchers had higher unadjusted and adjusted hazards of arm injury than relief pitchers (unadjusted HR = 2.0 [95% CI = 1.3, 3.0]; P = .002; adjusted HR = 2.4 [95% CI = 1.5, 4.0]; P < .001; Figure, Table 2).

Figure

Survival plot comparing starting and relief professional pitchers in days.

Figure

Survival plot comparing starting and relief professional pitchers in days.

Close modal
Table 2

Cox Proportional Hazards Model

Cox Proportional Hazards Model
Cox Proportional Hazards Model

Hazard of Elbow and Shoulder Injury Between Starting and Relief Pitchers

No differences were observed in the unadjusted or adjusted hazard of elbow injury between starting and relief pitchers (unadjusted HR = 1.7 [95% CI = 0.9, 3.2]; P = .101; adjusted HR = 1.9 [95% CI = 0.8, 4.2]; P = .130). Starting pitchers had higher unadjusted and adjusted hazards of shoulder injury than relief pitchers (unadjusted HR = 2.2 [95% CI = 1.3, 3.6]; P = .004; adjusted HR = 3.8 [95% CI = 2.0, 7.1]; P < .001).

Sensitivity Analyses

Starting pitchers had a greater adjusted hazard of arm injury involving 7+ (HR = 2.3 [95% CI = 1.3, 3.9]; P = .003) and 28+ (HR = 2.6 [95% CI = 1.5, 4.8]; P = .001) days lost to participation than relief pitchers. Compared with nonlinear analyses, Cox proportional hazards with assumed linearity for confounder results demonstrated a decreased hazard of arm injury, with starting pitchers incurring a greater hazard of arm injury than relief pitchers (HR = 2.1 [95% CI = 1.3, 3.5]; P = .002). Starting pitchers had greater odds of arm injury than relief pitchers based on logistic analyses for nonlinear (odds ratio = 2.5 [95% CI = 1.3, 4.9]; P = .008) and linear (odds ratio = 2.3 [95% CI = 1.2, 4.4]; P = .008) confounder calculations.

Our main findings were that after we controlled for confounders (holding other variables constant, including the number of pitches and appearances between starters and relievers), MiLB starting pitchers displayed a 2.4 times greater hazard of arm injury than MiLB relief pitchers. After separating elbow and shoulder injuries, we observed no difference for the hazard of elbow injury between MiLB starting and relief pitchers, but starting pitchers showed an almost 4 times greater hazard of shoulder injury.

Arm Injuries in Starting Versus Relief Pitchers

Starting pitchers had greater RR, RD, and hazard of arm injury than relief pitchers. Although arm injuries in starting and relief pitchers have not previously been directly compared, researchers10,19  noted differences between the pitching performances of MLB starters and relievers after injury. Relief pitchers exhibited decreased time to return to sport19  and improved pitching workload compared with starters.10  In both studies,10,19  the authors hypothesized that these differences between starters and relievers may be due to overall pitch load. However, we controlled for (held these confounders constant between starters and relievers) the time to event, total seasonal pitch load (number of pitches in a season), and number of pitching appearances. One possible explanation for these findings may be differences in pitch velocity. Greater MLB fastball pitch velocity has been associated with improved pitching performance26,27  but also an increased injury risk.26  At the MiLB level, pitchers throwing with consistently greater velocity are usually regarded as better prospects and are often placed in starting pitcher roles for increased pitching experience and development. It is not until the upper MiLB or MLB levels that these pitchers may be assigned to reliever roles, allowing for a large proportion of these faster-throwing pitchers to be regarded as starting pitchers. Another explanation may be the difference in induced cumulative pitching fatigue. Within a baseball game, the hardest pitches were thrown in the 1st and the 7th through 9th innings, and vertical movement and release height decreased from innings 2 through 6.27  On average, a starting pitcher throws between 5 and 7 innings, demonstrating the relationship of fatigue in the increased game workload from innings 2 through 6. Thus, the cumulative fatigue produced per pitching appearance may be greater in starters versus relievers, even after controlling for the seasonal pitch count and number of pitching appearances. As increased pitching fatigue is related to injury,28  the cumulative fatigue experienced by starting pitchers may increase their risk for arm injury compared with relievers. However, this is only speculation, as we did not measure fatigue. Future researchers should seek to understand how acute changes in pitch load and fatigue affect the risk of arm injury.

Elbow and Shoulder Injuries

The hazard of elbow injury was similar between starters and relievers, but after we adjusted for confounders, the former demonstrated a greater hazard of shoulder injury than the latter. Nonetheless, these findings should be interpreted with caution as the analyses had decreased power and the CIs were wide (ie, decreased precision of estimate). Elbow and shoulder injury risk factors were dissimilar in professional baseball pitchers.6  Specifically, differences in shoulder ROM may play a role in altering pitching biomechanics, increasing stress at the elbow or shoulder.29  For example, the late cocking phase induces the greatest elbow-valgus torque, which is produced by a combination of shoulder rotation, scapulothoracic extension, and trunk extension.29  In contrast, the greatest shoulder force is produced after ball release as the rotator cuff provides an eccentric contraction for arm deceleration.30  Investigators12  have identified differences in shoulder ROM between starting and relieving pitchers, with relievers displaying greater shoulder IR, ER, and TROM. However, in our study, shoulder ROM during spring training was similar between starting and relief pitchers. Other characteristics such as pitching biomechanics may bias starter and relief pitchers to different elbow or shoulder injury risk profiles.12  Clinicians should consider both elbow and shoulder injury predispositions when evaluating starter and relief pitchers. Furthermore, due to the disparities in shoulder injuries between groups, the number of physical examinations (including both shoulder ROM and strength) may need to increase for starters during the season compared with relievers.

Sensitivity Analyses

Seven- and 28-day arm injury and odds ratio analyses demonstrated similar results as the primary analyses. The RR and RD analyses also indicated an increased arm injury risk in starting pitchers versus relievers. However, assuming confounder linearity resulted in a decreased hazard ratio calculation. Within clinical medicine, many variables have a nonlinear relationship with a specific outcome. For example, BMI has a U-shaped relationship with mortality, as both low and high BMI values have an increased mortality risk.31  We observed that multiple confounders had nonlinear relationships with time to arm injury, including seasonal pitch load, number of pitching appearances, TROM, HA, and HT. As all continuous variables were evaluated for nonlinear transformations through a strict application of the Akaike Information Criteria, residuals, visualization, and conceptual biological plausibility, using these nonlinear transformations may have increased the precision of these models. The results suggest that assuming linearity in regard to continuous variables may underestimate the injury risk in professional pitchers; when developing injury risk models, potential nonlinear relationships should be assessed.

Strengths and Possible Limitations

We studied a 7-year prospective cohort specifically to investigate arm injuries in professional pitchers. Missing data were few, allowing for a complete case analysis and decreasing the risk of bias. Pitchers were recruited from all MiLB levels, increasing the generalizability of these findings. However, shoulder ROM can change between pitching appearances and throughout the season, decreasing the clinical utility of these findings.32  The internal response to the extrinsic pitching load can affect fatigue and ultimately the injury risk, decreasing the precision of these findings. We lacked access to participants' injury histories that did not involve orthopaedic surgery before they signed professional contracts. Furthermore, previous lower extremity and trunk injuries were not included. As previous injury can influence the injury risk, some residual confounding is present in these analyses. Also, out-of-game pitches and throws were not collected; thus, the total throwing-related load may be underestimated, decreasing the clinical utility of the results.

The MiLB starting pitchers demonstrated a 2.4 times greater hazard of arm injury than the MiLB relief pitchers. Subgroup analyses showed that MiLB starters exhibited a greater hazard of shoulder injury than relievers, but no differences were observed for the hazard of elbow injury. Still, due to the wide CIs, these subgroup analyses should be interpreted with caution. Clinicians may need to consider cumulative exposure and fatigue and how these factors relate to different pitching roles when assessing the injury risk for the pitching arm. Future researchers should investigate how acute changes in pitch load and fatigue affect the risk of arm injury.

1. 
Camp
CL,
Conte
S,
D'Angelo
J,
Fealy
SA.
Epidemiology of ulnar collateral ligament reconstruction in Major and Minor League Baseball pitchers: comprehensive report of 1429 cases
.
J Shoulder Elbow Surg
.
2018
;
27
(5)
:
871
878
.
2. 
Conte
S,
Requa
RK,
Garrick
JG.
Disability days in major league baseball
.
Am J Sports Med
.
2001
;
29
(4)
:
431
436
.
3. 
Conte
SA,
Fleisig
GS,
Dines
JS,
et al
Prevalence of ulnar collateral ligament surgery in professional baseball players
.
Am J Sports Med
.
2015
;
43
(7)
:
1764
1769
.
4. 
Knowles
SB,
Marshall
SW,
Miller
T,
et al
Cost of injuries from a prospective cohort study of North Carolina high school athletes
.
Inj Prevent
.
2007
;
13
(6)
:
416
421
.
5. 
Shanley
E,
Kissenberth
MJ,
Thigpen
CA,
et al
Preseason shoulder range of motion screening as a predictor of injury among youth and adolescent baseball pitchers
.
J Shoulder Elbow Surg
.
2015
;
24
(7)
:
1005
1013
.
6. 
Wilk
KE,
Macrina
LC,
Fleisig
GS,
et al
Deficits in glenohumeral passive range of motion increase risk of elbow injury in professional baseball pitchers: a prospective study
.
Am J Sports Med
.
2014
;
42
(9)
:
2075
2081
.
7. 
Noonan
TJ,
Thigpen
CA,
Bailey
LB,
et al
Humeral torsion as a risk factor for shoulder and elbow injury in professional baseball pitchers
.
Am J Sports Med
.
2016
;
44
(9)
:
2214
2219
.
8. 
Bradbury
JC,
Forman
SL.
The impact of pitch counts and days of rest on performance among major-league baseball pitchers
.
J Strength Cond Res
.
2012
;
26
(5)
:
1181
1187
.
9. 
Olsen
SJ
II,
Fleisig
GS,
Dun
S,
Loftice
J,
Andrews
JR.
Risk factors for shoulder and elbow injuries in adolescent baseball pitchers
.
Am J Sports Med
.
2006
;
34
(6)
:
905
912
.
10. 
Jones
KJ,
Conte
S,
Patterson
N,
ElAttrache
NS,
Dines
JS.
Functional outcomes following revision ulnar collateral ligament reconstruction in Major League Baseball pitchers
.
J Shoulder Elbow Surg
.
2013
;
22
(5)
:
642
646
.
11. 
Love
S,
Aytar
A,
Bush
H,
Uhl
TL.
Descriptive analysis of pitch volume in Southeastern Conference baseball pitchers
.
N Am J Sport Phys Ther
.
2010
;
5
(4)
:
194
200
.
12. 
Kibler
WB,
Sciascia
A,
Moore
S.
An acute throwing episode decreases shoulder internal rotation
.
Clin Orthop Rel Res
.
2012
;
470
(6)
:
1545
1551
.
13. 
Powell
JW,
Barber-Foss
KD.
Sex-related injury patterns among selected high school sports
.
Am J Sports Med
.
2000
;
28
(3)
:
385
391
.
14. 
Orchard
JW,
Kountouris
A,
Sims
K.
Incidence and prevalence of elite male cricket injuries using updated consensus definitions
.
Open Access J Sports Med
.
2016
;
7
:
187
194
.
15. 
Rauh
MJ,
Margherita
AJ,
Rice
SG,
Koepsell
TD,
Rivara
FP.
High school cross country running injuries: a longitudinal study
.
Clin J Sport Med
.
2000
;
10
(2)
:
110
116
.
16. 
Rae
K,
Orchard
J.
The Orchard Sports Injury Classification System (OSICS) version 10
.
Clin J Sport Med
.
2007
;
17
(3)
:
201
204
.
17. 
Orchard
JW,
Ranson
C,
Olivier
B,
et al
International consensus statement on injury surveillance in cricket: a 2016 update
.
Br J Sports Med
.
2016
;
50
(20)
:
1245
1251
.
18. 
Takenaga
T,
Goto
H,
Sugimoto
K,
et al
Left-handed skeletally mature baseball players have smaller humeral retroversion in the throwing arm than right-handed players
.
J Shoulder Elbow Surg
.
2017
;
26
(12)
:
2187
2192
.
19. 
Cerynik
DL,
Ewald
TJ,
Sastry
A,
Amin
NH,
Liao
JG,
Tom
JA.
Outcomes of isolated glenoid labral injuries in professional baseball pitchers
.
Clin J Sport Med
.
2008
;
18
(3)
:
255
258
.
20. 
Shanley
E,
Thigpen
C,
Clark
JC,
et al
Changes in passive range of motion and development of glenohumeral internal rotation deficit (GIRD) in the professional pitching shoulder between spring training in two consecutive years
.
J Shoulder Elbow Surg
.
2012
;
21
(11)
:
1605
1612
.
21. 
Shanley
E,
Rauh
MJ,
Michener
LA,
Ellenbecker
TS,
Garrison
JC,
Thigpen
CA.
Shoulder range of motion measures as risk factors for shoulder and elbow injuries in high school softball and baseball players
.
Am J Sports Med
.
2011
;
39
(9)
:
1997
2006
.
22. 
Retrosheet
.
The Baseball Cube
.
Accessed June 1,
2021
.
23. 
Sports Reference
.
Baseball Reference
.
Accessed June 1,
2021
.
24. 
Shanley
E,
Bailey
L,
Sandago
MP,
Pinkerton
A,
Singleton
SB,
Thigpen
CA.
The use of a pitch count estimator to calculate exposure in collegiate baseball pitchers
.
Phys Ther Sport
.
2015
;
16
(4)
:
344
348
.
25. 
Durrleman
S,
Simon
R.
Flexible regression models with cubic splines
.
Stats Med
.
1989
;
8
(5)
:
551
561
.
26. 
Bushnell
BD,
Anz
AW,
Noonan
TJ,
Torry
MR,
Hawkins
RJ.
Association of maximum pitch velocity and elbow injury in professional baseball pitchers
.
Am J Sports Med
.
2010
;
38
(4)
:
728
732
.
27. 
Whiteside
D,
Martini
DN,
Zernicke
RF,
Goulet
GC.
Ball speed and release consistency predict pitching success in Major League Baseball
.
J Strength Cond Res
.
2016
;
30
(7)
:
1787
1795
.
28. 
Erickson
BJ,
Sgori
T,
Chalmers
PN,
et al
The impact of fatigue on baseball pitching mechanics in adolescent male pitchers
.
Arthroscopy
2016
;
32
(5)
:
762
771
.
29. 
Fleisig
GS,
Andrews
JR,
Dillman
CJ,
Escamilla
RF.
Kinetics of baseball pitching with implications about injury mechanisms
.
Am J Sports Med
.
1995
;
23
(2)
:
233
239
.
30. 
Werner
SL,
Guido
JA
Jr,
Stewart
GW,
McNeice
RP,
VanDyke
T,
Jones
DG.
Relationships between throwing mechanics and shoulder distraction in collegiate baseball pitchers
.
J Shoulder Elbow Surg
.
2007
;
16
(1)
:
37
42
.
31. 
Troiano
RP,
Frongillo
EA
Jr,
Sobal
J,
Levitsky
DA.
The relationship between body weight and mortality: a quantitative analysis of combined information from existing studies
.
Int J Obese Rel Metab Disord
.
1996
;
20
(1)
:
63
75
.
32. 
Reinold
MM,
Wilk
KE,
Macrina
LC,
et al
Changes in shoulder and elbow passive range of motion after pitching in professional baseball players
.
Am J Sports Med
.
2008
;
36
(3)
:
523
527
.
Appendix

Knot Locations

Knot Locations
Knot Locations