Chronic Adaptation of the Coracohumeral Space and Subscapularis Tendon in Professional Baseball Pitchers

Abstract

Background:

Pitchers frequently experience anterior shoulder pain, possibly associated with coracohumeral impingement; however, whether the coracohumeral distance (CHD) and/or subscapularis tendon adapt chronically (bilateral difference) due to pitching, and whether clinical measures are associated with CHD and subscapularis tendon organization have not been evaluated in professional pitchers.

Hypothesis:

The authors hypothesized that dominant arm CHD would be smaller than the nondominant arm, dominant subscapularis tendon would have increased spatial frequency (ie, be more disorganized), and humeral retroversion (HR) would predict CHD and subscapularis tendon organization.

Level of Evidence:

Level 4.

Methods:

Healthy professional baseball pitchers were recruited during their preseason physical examination. Bilateral diagnostic ultrasound measured CHD, HR, and posterior capsule thickness (PCT), and quantified subscapularis tendon organization. External rotation, neutral, and crossbody CHD was measured.

Results:

Overall, 52 healthy professional baseball pitchers participated. The dominant arm of pitchers demonstrated a significantly narrower CHD in all 3 positions (P < 0.01), increased scapular protraction (163 vs 156 mm; P < 0.01), and increased spatial frequency of the subscapularis tendon (1.8 vs 1.6 peaks/mm; P < 0.01). HR was associated with CHD in 30° of external rotation (R² = 0.12; P < 0.01), neutral rotation (R² = 0.11; P < 0.01), and the crossbody position (R² = 0.28; P < 0.01). PCT was associated with CHD in 30° of external rotation (R² = 0.16; P = 0.05). HR and CHD in 30° of external rotation was associated most strongly with subscapularis tendon organization (R² = 0.11; P = 0.03).

Conclusion:

The dominant shoulder of professional pitchers presents with a smaller CHD, more scapular protraction, and more subscapularis tendon disorganization than the nondominant shoulder.

Clinical Relevance:

Professional pitchers demonstrate chronic CHD and subscapularis tendon adaptations, which may increase their risk for anterior shoulder pain and subscapularis tendon injury.

Over 8000 injuries occur per year in professional baseball players. ⁵ Anterior shoulder pain is a common source of pain and injury in baseball players, and determining the specific source of anterior shoulder pain is challenging. Commonly studied sources of anterior shoulder pain in baseball players include subacromial impingement, biceps tendon pathology, and/or labral pathology; however, subcoracoid impingement is relatively understudied in baseball. Subcoracoid impingement results from decreased coracohumeral distance (CHD), which is the distance between the coracoid process of the scapula and the lesser tubercle of the humerus. ¹⁹ Several studies have shown that, in the general population, a narrow CHD is associated with anterior shoulder pain related to a diagnosis of biceps brachii tendinopathy and/or subscapularis tears.^1,16,31

During the follow-through phase of pitching, the arm moves to the end range of shoulder adduction, internal rotation, and shoulder flexion in an effort to absorb the large eccentric forces. This shoulder position also creates the narrowest CHD and recreates symptoms in patients with anterior shoulder pain. ¹⁹ Also, several tissue adaptations at the shoulder that have been identified in pitchers that have the potential to further decrease the CHD space. For example, the dominant shoulder of pitchers has a thicker and stiffer posterior capsule compared with the nondominant arm,^27,30 with increased posterior capsule thickness (PCT) being related to increase soft tissue glenohumeral internal rotation deficit. ¹⁸ These changes in the posterior capsule may relate to CHD because tightening of the posterior capsule results in greater anterior translation of the humeral head on the glenoid fossa during glenohumeral flexion and horizontal adduction. ⁹ Posterior shoulder tightness has also been shown to increase forward scapular posture and scapular protraction in the dominant arm of pitchers. ¹⁰ Increasing forward shoulder posture could change the position of the coracoid process relative to the lesser tubercle of humerus, potentially narrowing the CHD. In addition, increased humeral retroversion (HR) in the dominant arms of pitchers results in the lesser tubercle of the humerus approximating closer to the coracoid process and decreasing the CHD.^6,7,22,24

As the CHD distance narrows, the subscapularis tendon will be subject to increased compressive forces by the coracoid process and lesser tubercle of the humerus. Previous research in rat supraspinatus tendons has established that when tendons are compressed they present a phenotype more closely resembling fibrocartilage. ² Degenerated supraspinatus and subscapularis tendons in human cadavers have shown differences in collagen structure compared with healthy tendons. ²⁶ In humans, diagnostic ultrasound has shown differences in collagen structure between degenerated and healthy supraspinatus and subscapularis tendons, and previous research has also associated greater collagen disorganization in the supraspinatus tendons of swimmers experiencing pain and disability.^25,28 Thus, the compressive force imposed by a decrease in CHD may lead to changes in the collagen organization of the subscapularis tendon, which may be related to anterior shoulder pain and other pathology in baseball pitchers. However, adaptations to the CHD and the subscapularis tendon have not been evaluated in professional pitchers.

Therefore, the primary purposes of this study were to examine the bilateral adaptation to the CHD in baseball pitchers and to determine whether the subscapularis tendon undergoes chronic adaptation due to pitching. This study will also determine whether glenohumeral and scapular adaptations (HR, PCT, and forward shoulder posture) are related to CHD and/or subscapularis spatial frequency (a measure of tendon organization). The authors hypothesized that the throwing arm will demonstrate a decreased CHD compared with the nonthrowing arm, that the dominant subscapularis tendon would have increased spatial frequency, and that CHD and subscapularis tendon organization will be related to glenohumeral and scapular adaptations.

Methods

Study Design and Participants

This cross-sectional study was approved with Institutional Review Board (Temple University) approval and participant informed consent. A total of 52 healthy professional baseball pitchers (mean age, 22.6 ± 2.2 years; height, 187.7 ± 5.8 cm; weight 92.7 ± 7.8 kg; 73% right-hand dominant) from a single organization volunteered for participation at the beginning of the 2021 Major League Baseball spring training. Pitchers were included only if they had >5 years pitching experience, were currently cleared by the medical team for full participation, did not have a shoulder surgery within the past year, and were between 18 and 35 years old. Participants were excluded if they had any shoulder surgery within the previous year or pitched >60 pitches in the previous 5 days in an effort to minimize the acute effects of pitching on ultrasound imaging. Diagnostic ultrasound was used bilaterally to measure CHD, HR, and PCT, and to quantify subscapularis tendon organization.

The handheld double square was used to measure scapular protraction as the wall to the anterior tip of the participants’ acromion process. Using the double square to measure scapular protraction has been found to have an intraclass correlation coefficient (ICC) of 0.89. ²⁰ Each participant was instructed to move backwards until their back and heels touched the wall. Then, the double square was positioned over the participant’s shoulder with 1 square flush against the wall. The second square was adjusted until it touched the tip of the participant’s acromion, and that value was recorded (in millimeters) as the subject’s scapular protraction. Three measurements were taken, and the mean was used for data analysis.

Ultrasound

The M-Turbo Ultrasound System Scanner (FujiFilm Sonosite Inc) with a 15-MHz linear-array transducer was used for all ultrasound assessments. Previous studies have found diagnostic ultrasound to have good reliability,^17,31 with excellent intrarater reliability in CHD assessment observed by the current study investigator, who had 2 years of experience performing this technique before data collection (Table 1; ICC > 0.99 in all positions). CHD was measured bilaterally with the shoulder in 3 positions: adducted with 30° of shoulder external rotation; neutral shoulder rotation with the forearm resting on the thigh; and with the hand placed on the opposite shoulder, placing the shoulder in internal rotation and horizontal adduction. The coracoid process was palpated and the transducer was placed over the anterior shoulder to visualize the head of the humerus and the tip of the coracoid process. The CHD was measured from the tip of the coracoid process to the closest aspect of the humeral head. Three images were captured in each position, saved, and exported for further analysis. The mean CHD from the 3 images was used for data analysis.

Table 1.

Intrarater reliability measures for dependent variables

Variable a	Mean ± SD	ICC	SEM
CHD external rotation, mm	15.8 ± 2.8	0.998	0.13
CHD neutral, mm	14.3 ± 2.6	0.991	0.24
CHD crossbody, mm	11.1 ± 1.8	0.992	0.17
Scapular protraction, mm	144.6 ± 19.9	0.997	1.10

CHD, coracohumeral distance; ICC, Intraclass correlation coefficient; SEM, standard error of measure.

^aThe 3 CHD measurements are performed with ultrasound, while scapular protraction is quantified with the double square.

To quantify HR bilaterally, subjects laid supine with the shoulder passively abducted to 90° and the elbow flexed passively to 90°. The investigator placed the transducer on the anterior shoulder, perpendicular to the long axis of the humerus. The investigator then rotated the humerus internally, centering the bicipital groove on the ultrasound image. The ultrasound was positioned so that a line connecting the greater and lesser tubercles was drawn parallel to the horizontal plane. Finally, a second investigator placed a digital inclinometer on the ulnar side of the forearm and recorded the forearm inclination angle. The measurement was completed 3 times and the mean HR angle was determined. Excellent reliability has been reported previously for quantifying HR. ¹⁵

PCT was assessed bilaterally with the participant seated and their forearm resting on the thigh. The ultrasound transducer was placed on the posterior shoulder to visualize the humeral head, rotator cuff, and glenoid labrum. The posterior capsule was identified as the structure directly lateral to the glenoid labrum, between the humeral head and rotator cuff. An image of the posterior capsule was captured, saved, and exported for further analysis. The study investigator assessing PCT has previously validated the use of ultrasound to measure PCT with a standard error of measurement (SEM) of 0.22 mm. ³⁰

Image Analysis

Subscapularis tendon structure was analyzed using custom MATLAB software (MathWorks) to determine peak spatial frequency. Using the ultrasound images taken in 30° of glenohumeral external rotation, the researcher identified the edge of the subscapularis tendon footprint on the lesser tubercle of the humerus and the myotendinous junction. A vertical line was placed on the most lateral aspect of the footprint and the medial edge of the myotendinous junction. A third line was then placed in the middle of the 2 existing lines. Last, 2 vertical lines were placed: 1 bisecting the middle line and medial line, and 1 bisecting the middle line and the lateral line (Figure 1).

Figure 1.

Ultrasound image of the subscapularis tendon in 30° of external rotation showing the locations of peak spatial frequency measurement (yellow lines).

A 1-dimensional (1-D) fast-Fourier transform was used to determine the spectral power of echogenicity over the length of each line. Peak spatial frequency was calculated for each of the 5 lines from the image of the subscapularis footprint and averaged for each tendon. The peak spatial frequency for an individual tendon is related to the spacing of the collagen fascicles averaged between the lateral edge of the subscapularis tendon footprint and the myotendinous junction.

Statistical Analysis

All statistical analysis was conducted using SPSS Version 26 (IBM). Paired t tests were utilized to compare the CHD and spatial frequency of the dominant and nondominant arms. A forward stepwise linear regression was used to examine the relationship between the clinical measures (HR, PCT, scapular protraction) and CHD (ER, neutral, cross-adduction). Another forward stepwise linear regression was used to examine the relationship between clinical measures (HR, PCT, scapular protraction), CHD (ER, neutral, cross-adduction), and subscapularis spatial frequency. Cohen’s d effect sizes were calculated for all bilateral comparisons of CHD. The P value was set to 0.05.

Results

The dominant shoulder had decreased CHD in all 3 positions tested (all P < 0.05) (Table 2). The dominant arm was also found to have significantly more scapular protraction (P < 0.01; effect size, 0.617) and a larger peak spatial frequency (P < 0.01; effect size, 0.21) compared with the nondominant arm.

Table 2.

Bilateral comparison of CHD, scapular protraction, and subscapularis tendon peak spatial frequency ^a

Variable	Dominant	Nondominant	Effect Size	P value
CHD external rotation, mm	16.2 (3.0)	17.1 (2.6)	0.461	<0.01*
CHD neutral, mm	14.2 (3.0)	14.9 (2.4)	0.406	<0.01*
CHD cross-body, mm	9.8 (2.8)	11.2 (2.4)	0.848	<0.01*
Scapular protraction, mm	162.6 (15.9)	156.3 (16.0)	0.617	<0.01*
Subscapularis PSF, peaks/mm	1.8 (0.2)	1.6 (0.1)	0.210	<0.01*

CHD, coracohumeral distance; PSF, peak spatial frequency.

^aData presented as mean (SD).

^*P < 0.05.

HR was found to be correlated significantly and negatively with CHD while the shoulder was in 30° of external rotation (R² = 0.12; P < 0.01), neutral rotation (R² = 0.11; P < 0.01), and in the cross-body position (R² = 0.28; P < 0.01) (Table 3). PCT was correlated significantly and positively with CHD in 30° of external rotation (R² = 0.16; P = 0.05), but was not predictive of CHD in the other 2 positions. Scapular protraction was not found to be predictive of CHD (all P > 0.05).

Table 3.

Stepwise regression analysis predicting CHD with the bilateral shoulders in 3 positions: external rotation, neutral, and cross-body

CHD: External Rotation
Predictor	R 2	SE	Beta	P
HR, deg	0.12	0.02	–0.35	<0.001*
PCT, mm	0.16	0.52	0.19	0.048*
Scapular protraction, mm			–0.04	0.479
CHD: Neutral
Predictor	R 2	SE	Beta	P
HR, deg	0.11	0.19	–0.32	0.001*
PCT, mm			0.18	0.056
Scapular protraction, mm			0.02	0.870
CHD: Cross-Body
Predictor	R 2	SE	Beta	P
HR, deg	0.28	0.02	–0.52	<0.01*
PCT, mm			–0.30	0.73
Scapular protraction, mm			0.10	0.23

CHD, coracohumeral distance; HR, humeral retroversion; PCT, posterior capsule thickness; SEM, standard error of measurement.

^*P < 0.05.

HR was the first model correlated significantly and positively with peak spatial frequency in the stepwise regression analysis (Table 4) (R² = 0.06; P = 0.02). The second model output from the regression analysis was a combination of HR and CHD in 30° of external rotation, which was also correlated significantly and positively with peak spatial frequency (R² = 0.11, P = 0.03). PCT, scapular protraction, CHD in the neutral rotation, and CHD in the crossbody position were not predictive of peak spatial frequency (all P > 0.05).

Table 4.

Stepwise regression analysis predicting subscapularis tendon peak spatial frequency bilaterally

Model	Predictor	Beta	R 2	SE	P
1	HR, deg	0.25	0.06	0.16	0.02*
2	HR, deg	0.33	0.11	0.16	0.03*
	CHD external rotation, mm	0.23
Excluded	CHD neutral, mm	0.12			0.29
Excluded	CHD crossbody, mm	0.12			0.31
Excluded	PCT, mm	0.15			0.14
Excluded	Scapular protraction, mm	0.13			0.20

CHD, coracohumeral distance; HR, humeral retroversion; PCT, posterior capsule thickness; SE, standard error of measurement.

^*P < 0.05.

Discussion

The hypotheses of the current study were supported as the throwing arm demonstrated a narrower CHD compared with the nonthrowing arm in all 3 positions, the dominant subscapularis tendon had increased spatial frequency, and both the CHD and the subscapularis tendon organization were related to shoulder adaptations (especially HR). Overall, it appears that the CHD on the dominant arm of pitchers adapts negatively to the high demands of pitching, potentially causing repetitive compression of the subscapularis tendon, which may lead to anterior shoulder pain.

Multiple studies in the general population have associated a narrower CHD with subscapularis tendon pathology.^3,16 Gerber et al ⁸ found that a CHD of <6 mm in an adduction and internal rotation position was associated with subscapularis tendon tears; furthermore, more recent research has suggested that a CHD of 7.6 mm in an adduction position had a sensitivity of 84.4% and specificity of 88.6% to predict subscapularis tendon pathology. ¹² Another study suggested a CHD of 9.5 mm in an adduction position has a sensitivity of 83.6% and a specificity of 83.9% to predict subscapularis tendon pathology. ²³ The mean dominant arm CHD of 9.8 mm in the crossbody position is approaching this 9.5 mm, supporting the view that the chronic adaptations in CHD in the throwing shoulder may increase the risk of anterior shoulder pain in pitchers. Lo and Burkhart ¹³ suggested that when the CHD decreases it has a “roller ringer” effect on the subscapularis tendon as it moves between internal and external rotation. This will create compression stress on the extra-articular side of the tendon and excessive tensile stress on the articular side. They defined this tear mechanism as tensile undersurface fiber failure (TUFF). ¹³ These proposed mechanical mechanisms of subscapular adaptations support both our CHD results (especially in the cross-adduction position) and the tendon organization results.

When examining the tendon organization results, we found that the dominant shoulder had increased spatial frequency (decreased organization). When tendons experience repetitive compressive forces they begin to demonstrate changes in normal tendon composition resulting in a phenotype more similar to cartilage, which has been shown to weaken the tendon and predispose it to pathology.^2,26 Coupled with the Lo and Burkhart TUFF mechanism of injury, the increased tensile stress on the articular side can also be a mechanism for tendon disorganization. ¹³ Interestingly, the current study found that the dominant subscapularis tendon had increased spatial frequency compared with the nondominant subscapularis tendon. A recent study used similar methods to compare the supraspinatus spatial frequency in master swimmers with pain and disability to those without pain and disability. ²⁹ The swimmers with pain and disability demonstrated increased peak spatial frequency of the supraspinatus tendon compared with asymptomatic swimmers. Interestingly, the dominant arm of the pitchers in the current study demonstrated similar subscapularis tendon spatial frequency (1.8 ± 0.2 peaks/mm) compared with those swimmers who reported pain and disability (1.8 ± 0.1 peaks/mm), while the nondominant arm of pitchers demonstrated similar spatial frequency (1.6 ± 0.1 peaks/mm) compared with the asymptomatic swimmers (1.7 ± 0.1 peaks/mm).

Examining the CHD and subscapularis tendon spatial frequency bilaterally is important to identify chronic adaptations, using the nondominant arm as an ideal nonthrowing “control.” It is also important to identify risk factors that contribute to a smaller CHD; therefore, a stepwise regression analysis was conducted to identify modifiable and nonmodifiable risk factors contributing to a narrower CHD. The stepwise regression analysis found that HR was predictive of a narrower CHD in all positions tested. While previous research has consistently demonstrated the dominant arms of pitchers exhibit increased HR,^7,24,27 this is the first study to analyze the relationship between CHD and HR. A study conducted by Leite et al ¹¹ found that increased HR as measured by magnetic resonance imaging was related progressively to the severity of subscapularis injury in symptomatic shoulders. ¹¹ Leite et al ¹¹ found the mean HR of -28.6° ± 19.5° in shoulders without subscapularis symptoms, -49° ± 9° in shoulders with subscapularis tendinopathy, and an HR of -51° ± 11° in shoulders with subscapularis tendon tears. Although direct comparisons in HR cannot be made between the current study and those reported by Leite et al. ¹¹ due to differences in HR measurement modality and technique, both studies suggest that increased HR contributes to a smaller CHD, which may increase compression of the subscapularis tendon and cause pathologic changes.

Increased forward shoulder posture is another adaptation to the pitching shoulder that may influence the CHD. Pitchers from the current study had increased forward shoulder posture on their throwing side, which is consistent with previous research ¹⁰ ; however, increased forward shoulder posture did not relate to CHD. Instead HR is related to CHD in all positions, whereas PCT is also related to CHD, specifically in the externally rotated position. During the late cocking phase of the throw, pitchers’ shoulders are abducted horizontally and rotated externally. In this position, the subscapularis tendon wraps around the coracoid process in an “L” shape. This “L” shape creates a stressed position, in which the subscapularis tendon may undergo local changes in collagen deposition as type II collagen deposition increases with compressive stress (causing a transition from tendinous to cartilaginous tissue).^4,14 With the arm cocking position during pitching already placing the subscapularis in a stressed “L” position, increased HR also decreases the CHD, potentially adding a point where the “roller ringer” effect may take place, potentially explaining how HR relates to subscapularis tendon organization. Although not measured in the current study, a loss of external rotation range of motion has been observed in baseball pitchers and may be explained by a tight subscapularis musculotendinous unit. ²¹ If subscapularis tightness is present, it may exacerbate these effects on tendon disorganization. PCT was also found to increase CHD weakly in the externally rotated position. Previous research has demonstrated that a tight posterior capsule will move the humeral head posterior and superior when in an externally rotated position. However, since this was found to be a weak relationship and was observed only in the externally rotated position, it is unclear if this is protective of the subscapularis tendon.

Overall, clinicians should know that the throwing shoulder of professional pitchers undergoes chronic adaptations to CHD that may be related to negative subscapularis tendon changes. However, further research comparing pitchers with versus without anterior shoulder pain will be important in clarifying the role of CHD and subscapularis tendon adaptations in the onset of anterior shoulder pain. Also, pitchers with increased HR may be at a greater risk of anterior shoulder pain since HR is correlated negatively with CHD in all tested positions. These changes may lead to maladaptive chronic changes in the subscapularis tendon, as HR and CHD were both associated with changes to subscapularis spatial frequency. Better understanding the risk factors for a narrower CHD and whether the subscapularis tendon undergoes adaptations due to these risk factors can help clinicians and researchers minimize anterior shoulder pain in baseball pitchers.

This study has significant limitations. First, the participants were all healthy pitchers. In addition, this study was cross-sectional; tracking these adaptations over the course of a single full season would likely identify how the CHD and subscapularis tendon change over time. Additionally, this study recruited only professional pitchers, so the results may not be generalizable to youth or collegiate athletes. Finally, reproducing the study methodology would be limited as it requires the researchers to be skilled in the use of diagnostic ultrasound, and custom software was used to assess tendon organization.

Conclusion

The dominant shoulder of professional pitchers presents with a smaller CHD, more scapular protraction, and more subscapularis tendon disorganization (increased spatial frequency) than the nondominant shoulder. HR is associated with CHD in all 3 arm positions, and the combination of increased HR and increased PCT is associated with subscapularis tendon organization. Overall, professional pitchers demonstrate chronic adaptations of their CHD and subscapularis tendon that may increase their risk for anterior shoulder pain and subscapularis tendon injury.