The Relationship Between Maximum Shoulder Horizontal Abduction and Adduction on Peak Shoulder Kinetics in Professional Pitchers

Abstract

Background:

Repetitive horizontal shoulder abduction during pitching can cause increased contact between the posterosuperior aspect of the glenoid and the greater tuberosity of the humeral head, theoretically putting baseball pitchers at increased risk of shoulder internal impingement and other shoulder pathologies.

Hypothesis:

Increased shoulder horizontal abduction is associated with increased shoulder anterior force, while increased horizontal adduction is associated with increased shoulder distraction force.

Study Design:

Descriptive laboratory study.

Level of Evidence:

Level 4.

Methods:

A total of 339 professional baseball pitchers threw 8 to 10 fastball pitches using 3D motion capture (480 Hz). Pitchers were divided into 2 sets of quartiles based on maximum shoulder horizontal abduction and adduction. Elbow flexion, shoulder external rotation, and peak shoulder kinetics were compared between quartiles with post hoc linear regressions conducted for the entire cohort.

Results:

At maximum shoulder horizontal abduction, there was no difference in ball velocity between quartiles (P = 0.76). For every 10º increase in maximum shoulder horizontal abduction, shoulder anterior force decreased by 2.2% body weight (BW) (P < 0.01, B = −0.22, β = −0.38), shoulder adduction torque decreased by 0.5%BW × body height (BH) (P < 0.01, B = −0.05, β = −0.19), and shoulder horizontal adduction torque decreased by 0.4%BW × BH (P < 0.01, B = −0.04, β = −0.48). For every 10º increase in maximum shoulder horizontal adduction, shoulder anterior force increased by 2%BW and ball velocity decreased by 1.2 m/s (2.7 MPH).

Conclusion:

Professional pitchers with the least amount of maximum horizontal adduction had faster ball velocity and decreased shoulder anterior force. Pitchers with greater maximum shoulder horizontal abduction had decreased shoulder anterior force, shoulder adduction torque, and shoulder horizontal adduction torque. To maximize ball velocity as a performance metric while minimizing shoulder anterior force, pitchers can consider decreasing maximum shoulder adduction angles at later stages of the pitch.

Clinical Relevance:

Identifying risk factors for increased throwing shoulder kinetics (ie, shoulder anterior force, shoulder adduction torque) has potential implications in injury prevention. Specifically, mitigating shoulder anterior forces may be beneficial in reducing risk of injury.

The upper extremity kinematics associated with pitching a baseball contribute to increased risk of shoulder joint pathologies.^8,28 At the shoulder, abduction, horizontal abduction, and external rotation in baseball pitchers have been implicated in the increased risk of shoulder injury.^{2,6,8,12,15,20,25–27} In particular, increased horizontal abduction has negative biomechanical effects on the shoulder joint, ⁶ noted in multiple arthroscopic and cadaveric reports, with study subjects ranging from youth to adults.^2,12,20,27

Several kinetic variables are associated with shoulder horizontal abduction and adduction, many of which have been performed on adolescents,^18,19,26 or have included a mixture of both adolescents and adults.^7,25 Studies of high school Japanese baseball pitchers independently conducted by Takagi et al ²⁵ and Tanaka et al ²⁶ reported increased anterior shear force on the shoulder when players are in a position with increased horizontal abduction at ball release. Minimizing the amount of horizontal adduction at ball release similarly reduced the force experienced by the pitcher’s shoulder, in particular, distraction force that causes a separation between the arm and the glenohumeral joint. ²⁸ This distraction force has many implications on glenohumeral joint pathology and is associated with glenoid labrum and rotator cuff injuries. ²⁸

It is important to assess the forces and torques imparted on the throwing arm in professional pitchers, given previous studies have shown upper extremity kinetics of pitching movements increase with competition level.^7,25 Because of the increased kinetics (including shoulder distraction force and elbow varus torque) seen in higher competition levels, ¹⁴ these players may be at a higher risk of experiencing shoulder injury.^7,25 Despite being a population that is routinely afflicted by various shoulder pathologies, there remains a paucity of data on the mechanics of professional baseball players. To date, no report has focused on the implications of shoulder horizontal abduction or adduction angles in professional pitchers, in particular American pitchers, which have key kinematic differences from players in eastern (ie, Japanese, Korean) countries.^4,5,16,17

Previous groups have studied kinematic-kinetic relationships at foot contact, maximum shoulder external rotation, and at ball release, while none have explored at maximum shoulder horizontal adduction or abduction.^19,22,23 Maximum shoulder horizontal abduction is achieved at early portions of the pitch, while maximum shoulder horizontal adduction occurs later, near the time of ball release. The degree of abduction and adduction at these critical time points has yet to be examined as potential correlates with throwing arm kinetics. Therefore, the purpose of this study was to examine the effect of maximum shoulder horizontal abduction and maximum shoulder horizontal adduction on peak shoulder kinetics in professional pitchers. We hypothesized increased shoulder horizontal abduction is associated with increased shoulder anterior force, while increased horizontal adduction is associated with increased shoulder distraction force.

Methods

Motus Global (Rockville Center, NY) deidentified all data predistribution, qualifying for exempt review under federal guidelines approved by the Hospital for Special Surgery institutional review board (New York, NY). Inclusion requirements for participation were as follows: at the time of testing, pitchers were currently on the Major League or Minor League (low A, high A, AA, and AAA) roster and pitchers had no record of serious injury (requiring >2 weeks of rest or rehabilitation) in the past 6 months.

Pitching evaluations were conducted as previously discussed^4,11; pitchers reported to the test site where a privacy waiver was administered and consent was obtained. Demographic data were reported by the pitcher including age, preferred throwing arm, experience, and history of injury. Researchers measured and recorded the pitchers’ height and weight. The pitcher was given unlimited time to warm up with his preferred routine for pitching at maximum effort (ie, arm bands, stretching, plyometric care, long toss, etc). Once the pitcher indicated that he was ready to pitch, 46 reflective markers were placed on anatomic landmarks, pitchers were allowed to throw and get used to wearing the markers after marker placement. ¹¹ The 8-camera Raptor-E motion analysis system (Motion Analysis Corp) was used to record marker data (480 Hz). The global coordination system was established on the basis of the International Society of Biomechanics standards: Y was vertically upward, X was perpendicular to Y (positive to home plate), and Z was the cross product of X and Y.

Pitchers were instructed to pitch between 8 and 12 fastballs with game-like effort to a catcher behind home plate at regulation distance (18.4 m). Pitchers were allowed to pitch at their own set rate, and given the option to pitch from the stretch or the wind up. Ball velocity was collected with a radar gun located behind the pitcher (Stalker Sports Radar).

All data processing for building full body kinematics and throwing arm kinetics was performed in custom built Matlab scripts (The Mathworks) as described previously.^4,11 Data from the markers were filtered by a low pass filter (fourth-order, zero lag Butterworth filter, 13.4 Hz cutoff frequency). ¹⁰ Horizontal adduction and abduction of the shoulder was defined as the angle between the upper throwing arm and the upper trunk vector in the transverse plane of the upper trunk. The zero or neutral point was defined when the shoulder was laterally abducted to 90° (Figure 1). When the shoulder moved anterior to the zero point this defined adduction, when the shoulder moved posterior to the zero position this was defined as abduction. Maximum and minimum horizontal adduction were defined within the time points of foot contact and ball release.^4,11 The pitch was divided into 5 time points: foot contact, maximum shoulder horizontal abduction, maximum shoulder external rotation, maximum shoulder horizontal adduction, and ball release. Pitch timing was recalculated when ball release represented t = 0, with any events occurring before it represented as negative values. Foot contact was identified as the first frame when the lead toe or heel in the Z axis reached the minimum. Maximum shoulder external rotation was established during the frame in which the throwing arm reached maximum external rotation. Ball release was calculated as the instant 0.01 s after the wrist passed the elbow in the forward direction.^4,11 Peak kinetic values were calculated for shoulder internal rotation torque, shoulder distractive force, shoulder adduction torque, shoulder anterior shear force, shoulder superior force, and shoulder horizontal adduction torque. These kinetics were normalized by pitcher weight for force and the product of player weight and height for torque.^4,11

Figure 1.

Shoulder horizontal adduction and abduction throughout the entire pitch for professional pitchers.

Max H Abd, maximum horizontal abduction; Max H Add; maximum horizontal adduction; MER, maximum external rotation; Max H Add.

Pitches were averaged per player for all calculations and displayed as means and standard deviations. Pitchers were subdivided into quartiles based on maximum shoulder horizontal adduction and maximum shoulder horizontal abduction. Pitchers with the smallest maximum abduction/adduction quartiles were categorized as Q1, whereas pitchers with the largest maximum abduction/adduction were categorized as Q4. Analysis of variance was used for comparison between the 4 quartiles of maximum shoulder horizontal abduction and maximum shoulder horizontal adduction with ball velocity, shoulder external rotation, elbow flexion, and 6 kinetic shoulder variables as outcomes of interest. Post hoc analysis for variables that derived significance included (1) two-sample t test with Bonferroni correction and (2) linear regression correlation coefficients to quantify the amount of change shoulder horizontal abduction and adduction coincided with the outcomes of interest. Significance was set at P < 0.05. Matlab version R2020a (The Mathworks Inc) was used in all statistical analyses.

Results

A total of 339 professional pitchers was included in this study (21.9 ± 2.1 years; 189.7 ± 5.7 cm; 94.8 ± 9.5 kg). At foot contact, pitchers’ shoulders were horizontally abducted (38 ± 12°). The average maximal shoulder horizontal abduction (44 ± 13°) was achieved after foot contact. Pitchers then moved the shoulder into horizontal adduction at maximum external rotation (8 ± 8°), reaching maximal adduction (10 ± 9°). At the time of ball release, shoulders were in a neutral position (2 ± 9°) (seeFigure 1).

When pitchers were subdivided into quartiles based on maximum shoulder horizontal abduction (Table 1), shoulder adduction torque (P < 0.01), shoulder horizontal adduction torque (P < 0.01), and shoulder anterior force (P < 0.01) decreased with increasing horizontal abduction. Shoulder external rotation at maximum shoulder horizontal abduction decreased with each quartile (36 ± 29°, 24 ± 26°, 17 ± 27°, -2 ± 29°; P < 0.01).

Table 1.

Maximum shoulder horizontal abduction quartiles

	Q1 (n = 85)	Q2 (n = 85)	Q3 (n = 84)	Q4 (n = 85)	Significance
Kinematics
Maximum shoulder horizontal abduction, deg	28 ± 5	39 ± 2	47 ± 3	61 ± 8	a, b, c, d, e, f
Shoulder external rotation at MHAb, deg	36 ± 29	24 ± 26	17 ± 27	-2 ± 29	a, b, c, e, f
Elbow flexion at MHAb, deg	89 ± 24	83 ± 24	88.6 ± 22	79 ± 27
Ball velocity, m/s	38.3 ± 1.8	37.9 ± 2.2	38.3 ± 1.7	38.3 ± 2.0
Peak shoulder kinetics
Shoulder internal rotation torque, %BW × BH	5.0 ± 0.7	5.0 ± 0.7	5.2 ± 0.8	4.9 ± 0.8
Shoulder horizontal adduction torque, %BW × BH	6.2 ± 0.8	5.9 ± 0.8	5.6 ± 0.9	4.8 ± 0.9	a, b, c, e
Shoulder superior force, %BW	19.9 ± 8.3	16.5 ± 7.6	17.5 ± 9.6	17.4 ± 9.5
Shoulder anterior force, %BW	46.1 ± 7.3	43.3 ± 6.7	41.2 ± 6.3	38.0 ± 7.1	a, b, c, e, f
Shoulder adduction torque, %BW × BH	8.7 ± 1.9	8.7 ± 2.0	8.3 ± 2.4	7.1 ± 2.1	a, b, c
Shoulder proximal force, %BW	114.5 ± 14.9	113.6 ± 15.4	116.3 ± 16.7	117.7 ± 16.3

BH, body height; BW, body weight; MHAd, maximum shoulder horizontal abduction.

Values are presented as mean ± standard deviation.

Significant differences (P < 0.05) between (a) Q1 and Q2, (b) Q1 and Q3, (c) Q1 and Q4, (d) Q2 and Q3, (e) Q2 and Q4, (f) Q3 and Q4.

When pitchers were subdivided into quartiles based on maximum shoulder horizontal adduction (Table 2), Q4 had significantly slower ball velocity compared with Q1 and Q2 (37.7 ± 1.9 vs 38.6 ± 1.9, 38.5 ± 1.7 m/s; P = 0.03). Q1 had decreased shoulder anterior force compared with Q2, Q3, and Q4 (39.2 ± 7.1 vs 42.2 ± 6.5, 42.2 ± 7.2, 45.0 ± 8.0% body weight (BW); P < 0.01).

Table 2.

Maximum shoulder horizontal adduction quartiles

	Q1 (n = 85)	Q2 (n = 84)	Q3 (n = 85)	Q4 (n = 85)	Significance
Kinematics
Maximum shoulder horizontal adduction, deg	0 ± 4	7 ± 2	13 ± 2	22 ± 6	a, b, c, d, e, f
Shoulder external rotation at MHAd, deg	160 ± 16	156 ± 14	156 ± 19	160 ± 8
Elbow flexion at MHAd, deg	75 ± 14	77 ± 15	74 ± 15	81 ± 13
Ball velocity, m/s	38.6 ±1.9	38.5 ± 1.7	38.1 ± 2.1	37.7 ± 1.9	c, e
Peak shoulder kinetics
Shoulder internal rotation torque, %BW × BH	5.0 ± 0.7	5.2 ± 0.9	4.9 ± 0.7	5.0 ± 0.8
Shoulder horizontal adduction torque, %BW × BH	5.4 ± 1.2	5.6 ± 0.9	5.67 ± 1.0	5.8 ± 0.9
Shoulder superior force, %BW	18.5 ± 8.7	18.2 ± 10.2	16.6 ± 8.3	17.9 ± 8.1
Shoulder anterior force, %BW	39.2 ± 7.1	42.2 ± 6.5	42.2 ± 7.2	45.0 ± 8.0	a, b, c
Shoulder adduction torque, %BW × BH	8.0 ± 2.0	8.2 ± 2.1	8.5 ± 2.5	8.1 ± 2.0
Shoulder proximal force, %BW	114.8 ± 15.1	115.8 ± 14.4	116.1 ± 18.5	115.5 ± 15.2

BH, body height; BW, body weight; MHAd, maximum shoulder horizontal adduction.

Values are presented as mean ± standard deviation.

Significant differences (P < 0.05) between (a) Q1 and Q2, (b) Q1 and Q3, (c) Q1 and Q4, (d) Q2 and Q3, (e) Q2 and Q4, (f) Q3 and Q4.

Kinetic values and ball velocity that derived significance from quartile subdivisions for shoulder horizontal adduction were analyzed with regression correlation coefficients (Table 3). There was no significant relationship between maximum shoulder horizontal abduction and ball velocity (P = 0.49). For every 10° increase in maximum shoulder horizontal abduction, shoulder anterior force decreased by 2.2%BW (P < 0.01, B = −0.22, β = −0.38), shoulder adduction torque decreased by 0.5%BW × body height (BH) (P < 0.01, B = −0.05, β = −0.19), and shoulder horizontal adduction torque decreased by 0.4%BW × BH (P < 0.01, B = −0.04, β = −0.48). There was no significant relationship between the timing of maximum shoulder horizontal abduction and kinetics and ball velocity (P_min = 0.08). Significant relationships with maximum shoulder horizontal adduction and shoulder anterior force (P < 0.01, B = -8.62, β = 0.–0.08) and ball velocity (P < 0.01, B = −0.04, β = 0.12) were appreciated. For every 10º increase in maximum shoulder horizontal adduction, shoulder anterior force increased by 2%BW and ball velocity decreased by 1.2 m/s (2.7 MPH). For every 10 ms delay in achieving maximum horizontal adduction, shoulder anterior force decreased by 1.3%BW (P = 0.02, B = −0.13, β = −0.13) and shoulder horizontal adduction torque increased by 0.3%BW × BH (P < 0.01, B = 0.03, β = 0.24).

Table 3.

Shoulder horizontal abduction/adduction correlation coefficients with variables of interest

	Shoulder Anterior Force		Shoulder Adduction Torque		Shoulder Horizontal Adduction Torque		Ball Velocity
	B	β	B	β	B	β	B	β
Maximum shoulder horizontal abduction	−0.22	−0.38*	−0.05	−0.19*	−0.04	−0.48*	0.00	0.04
Maximum shoulder horizontal adduction	0.20	0.25*	−0.01	−0.02	0.01	0.08	−0.04	−0.12*
Timing of maximum horizontal abduction	−0.01	−0.08	0.00	0.05	0.00	0.10	0.00	0.03
Timing of maximum horizontal adduction	−0.13	−0.13*	0.00	0.01	0.03	0.24*	−0.01	−0.03

^*Statistical significance at P < 0.05.

Discussion

Pitchers who achieved greater maximal shoulder horizontal abduction (>60°, occurring close to foot contact) experienced decreased peak shoulder kinetics, while pitchers with increased shoulder horizontal adduction experienced increased shoulder anterior force with slower ball velocity. These results are contrary to our hypothesis that increasing maximum horizontal adduction or abduction would be positively associated with higher throwing shoulder kinetics suggesting maximizing the extremities of inpitch shoulder motions may not necessarily always have negative clinical implications.

Shoulder horizontal abduction at foot contact has been reported between 17° and 30° in professional pitchers.^3,24 At maximum external rotation and ball release, the literature has ranged from 5° to 14° at maximum external rotation, and 0° to 12° at ball release, comparable with what this study observed.^4,21 Progression through the pitch includes forward arm movement relative to the trunk with increased horizontal adduction, acted on by the anterior deltoid and pectoralis major. Increased trunk flexion and rotation causes the upper extremity to lag behind the trunk in a posterior position relative to the scapular plane. ¹ In response to rapid internal rotation of the humerus immediately after reaching maximum external rotation, the arm moves back toward a position of horizontal abduction and continues to near-zero values at ball release. ³

Increased maximum horizontal adduction of the shoulder posterior to the scapular plane is potentially deleterious to the anterior capsular structures. ¹ Results of this study suggest minimization of horizontal adduction toward the end of the pitch may decrease shoulder anterior forces most significantly, given Q1 had the smallest anterior force with approximately near-zero abduction/adduction. Takagi et al ²⁵ similarly found that the anterior and posterior forces at the point of maximum external rotation could be minimized by having a horizontal adduction angle of 1.8°. When the plane of the shoulder is directly in line with the upper torso, the glenohumeral joint is fixated in its most stable position and the least kinetics are applied. Maintaining this position in later stages of the pitch, therefore, may be favorable.

The prevalence of anterior shoulder pain associated with pitching has been well identified and documented. ¹⁵ One mechanism through which increased horizontal adduction poses potential risks for shoulder joint injury is internal impingement.^6,12 Repetitive excessive horizontal adduction of pitching can cause increased contact between the posterosuperior aspect of the glenoid and the greater tuberosity of the humeral head.^8,12 This repetitive movement also affects the supraspinatus and infraspinatus tendons because it increases the amount of contact between these tendons and the glenoid. ¹² Through this biomechanical mechanism, impingement not only increases a pitcher’s risk of developing shoulder pain but also can be a segue for additional injury.^2,9,20,27 Two commonly seen pathologies in pitchers that can be attributed to impingement are superior labrum anterior posterior tears^9,12and articular-sided rotator cuff tears.^12,20

Ball velocity negatively correlated with maximum shoulder horizontal adduction. Increased shoulder horizontal abduction and decreased shoulder horizontal adduction have both correlated with increased ball velocity; however, these were reported at foot contact and not maximum external rotation or ball release. ²⁴ Still, the prospect of attaining increased ball velocity with minimized horizontal adduction throughout the pitch can be substantiated. When pitchers begin to rotate their upper trunk, the humerus has to overcome the degree of horizontal abduction a pitcher starts with, which increases the time it takes to reach maximum horizontal adduction. ²⁴ With greater shoulder horizontal abduction and shoulder external rotation, the pitcher is able to apply an accelerating force to the ball for the longest distance, hence the reason why the cocking process takes 77% to 80% of the pitch, ultimately generating increased ball velocity as a result.^13,21 Excessively adducting the shoulder horizontally, or leading with the elbow, will in effect cause maximal shoulder horizontal adduction to be reached earlier, with less accelerant force placed on the ball.

Pitchers with greater maximum shoulder horizontal abduction had decreased throwing arm kinetics (shoulder horizontal adduction torque, shoulder adduction torque, shoulder anterior force) with no difference in ball velocity. This is a controversial finding as excessive horizontal abduction has previously been theorized to be deleterious to pitchers. ⁶ Takagi et al ²⁵ also noted increased shoulder anterior force in a mixed cohort of pitchers, with increased shoulder horizontal abduction at the moment of maximum shoulder external rotation, contrary to what we observed. These differences may be attributable to differences in pitching levels, as the current study investigated exclusively professional pitchers and Takagi et al ²⁵ used a mixed cohort. In addition, Takagi et al ²⁵ evaluated shoulder horizontal abduction at a specific time point in the pitch as opposed to at its maximum value. The current study investigated peak kinetics of throwing arm kinetics compared with using the value at a defined moment of interest (ie, maximum shoulder external rotation). Ultimately, additional evaluation is warranted to further characterize the role of shoulder horizontal abduction and its potential deleterious or nondeleterious effect on the throwing arm.

There were significant limitations to this study. Only professional pitchers were used in this analysis and the findings might not be applicable to pitchers of other skill levels. Additionally, while the kinetic values were calculated and the peaks for each value were used in this analysis, these peaks occurred during different phases of the pitch (arm cocking, arm acceleration, and arm deceleration), and may not adequately represent the kinetics experienced directly at maximum shoulder abduction/adduction as conducted by previous research groups.^25,26 Shoulder horizontal adduction increased beyond ball release (see Figure 1), suggesting the maximum shoulder horizontal adduction evaluated in the current study represented the “mid pitch” peak rather than the absolute peak value. Evaluating shoulder horizontal adductions’ influence on shoulder kinetics during the deceleration phase is a limitation of the current analysis. Finally, the regression coefficients established here are applied only at specific time points and only apply within a reasonable range. In particular, because only peak, positive kinetic values were included in analysis, how negative peak kinetic values (ie, shoulder posterior force, shoulder inferior force, etc) change with shoulder horizontal abduction is not clear and should not be extrapolated from these results.

Conclusion

Professional pitchers with the least amount of maximum shoulder horizontal adduction had faster ball velocity as well as decreased shoulder anterior force. Pitchers with the least amount of horizontal abduction had increased shoulder anterior force, shoulder adduction torque, and shoulder horizontal adduction torque. Pitchers with greater maximum shoulder horizontal abduction had decreased throwing arm kinetics. To maximize ball velocity as a performance metric outcome while minimizing shoulder anterior force, professional baseball players should consider decreasing shoulder adduction at later stages of the pitch.