Shoulder Range of Motion Deficits in Youth Throwers Presenting With Elbow Pain

Abstract

Background:

Glenohumeral internal rotation deficit (GIRD) and total arc of motion difference (TAMD) have been associated with elbow injuries in throwing athletes.

Hypothesis:

Youth pitchers with elbow pain will have greater GIRD and TAMD compared with youth pitchers without elbow pain.

Study Design:

Cross-sectional study.

Level of Evidence:

Level 3.

Methods:

Glenohumeral range of motion of 25 consecutive throwing athletes presenting with elbow pain and that of a matched control group of 18 asymptomatic throwing athletes were compared. Bilateral glenohumeral internal rotation, external rotation, and horizontal adduction at 90° were measured and GIRD and TAMD were then calculated. An analysis of variance was performed to compare range of motion between throwers with and without elbow pain.

Results:

The average GIRD of the elbow pain group was 32.7° compared with 14.5° in the control group (P < 0.05). The average TAMD in the elbow pain group was 28.3° compared with 6.7° in the control group (P < 0.05). GIRD and TAMD were present in 88% (22 of 25) and 96% (24 of 25) of the elbow pain group versus 33.3% (6 of 18) and 55.6% (10 of 18) of the control group, respectively.

Conclusion:

Compared with asymptomatic youth pitchers, those presenting with elbow pain have a statistically significant GIRD and TAMD.

Clinical Relevance:

This study suggests that a GIRD and TAMD may predispose youth pitchers to present with symptomatic elbow pain.

In recent years, there has been a significant increase in elbow pathology in adolescent throwers, which has been attributed to high pitch counts and year-round baseball schedules. ⁶ Diagnoses such as medial elbow apophysitis, ulnar collateral ligament injuries, osteochondritis dissecans of the elbow, and valgus extension overload injuries have become common in adolescent pitchers, with a 343% increase in ulnar collateral ligament reconstructions with a disproportionate trend in average annual incidence for patients between 15 and 19 years old. ¹² Because of the increase in elbow pathology in adolescent throwers, there is a significant need to identify potential modifiable physical risk factors that may be contributing to injury. Repetitive stress at the shoulder during throwing and pitching can lead to altered range of motion (ROM) patterns. A side-to-side loss of internal rotation (IR) found in the throwing shoulder, referred to as glenohumeral IR deficit (GIRD), may contribute to shoulder and elbow injuries.^{2-4,9,10,13,14,16,17,19-21,23,24} High school baseball players with GIRD ≥25° have a 4 to 5 times higher risk of upper extremity injury than players with GIRD <25°. ¹⁹

In addition to GIRD, total arc of motion (sum of shoulder IR and external rotation [ER]) is also important to assess in throwers. Total arc of motion differences (TAMD) between the throwing and nonthrowing shoulder that is greater than 5° may result in an increased risk of injury. ²³ In adult throwers, GIRD as well as TAMD have been associated with articular-sided rotator cuff injuries, superior labrum anterior to posterior (SLAP) tears, biceps tendon injuries, capsular injuries, and ulnar collateral ligament injuries.^2,7 These changes in ROM are multifactorial in the development of GIRD and TAMD.

While adolescent and older pitchers often have a rotational osseous adaptation (humeral head retroversion) in their shoulders, Nakamizo et al ¹⁴ found GIRD in youth throwers, who may have not yet developed these adaptive changes. In isolation, it is not a true injury, but rather an adaptation that may place the thrower at increased risk for shoulder and elbow injuries, while fewer shoulder problems have been observed with posterior capsule stretching.^1,10,21,23

However, no study has examined GIRD and TAMD in adolescent pitchers with elbow pathology. The purpose of this study is to evaluate GIRD and TAMD in adolescent pitchers with elbow pain compared with an asymptomatic age-, activity-, and position-matched cohort. We hypothesized that youth pitchers with elbow pain would have greater GIRD and TAMD than asymptomatic pitchers.

Methods

Participants

Institutional review board approval for the study was obtained. Prior to data collection, all testing procedures were explained to each participant and his or her parent and informed consent was obtained. Participants completed a questionnaire that included questions regarding height, weight, hand dominance, elbow pain, and throwing history. Exclusion criteria included prior throwing shoulder or elbow surgery, nonthrowing elbow trauma, or any concordant throwing shoulder pain within the past 2 years. Baseball pitchers between the ages of 11 and 18 years who presented to an orthopaedic clinic with a chief complaint of elbow pain were recruited to participate. A total of 25 patients with elbow pain were enrolled in the study. The control group consisted of 18 asymptomatic participants who were age and position matched to the group with elbow pain. Participant demographics, as well as clinical diagnosis of the elbow pain group, are presented in Table 1.

Table 1.

Demographics

	Control Group	Elbow Pain Group	P
Number	18	25
Age, y	14.6 (2.0)	14.8 (2.0)	0.693
Height, cm	172.8 (14.3)	174.4 (11.8)	0.689
Weight, kg	64.9 ± 15.8	71.2 ± 16.7	0.212
Level of play, n
Youth	7	6
High school	11	19
Average years pitched	5	6
Diagnosis
Ligament rupture		10
Tendinitis		7
Fracture		4
Ligament sprain		2
Muscle strain		2

Data Collection

After completion of the questionnaire, patients underwent bilateral shoulder ROM measurements using a validated Apple iPhone goniometer application and technique. ²² Measurements of both shoulders were taken with the patient supine on an examination table. Maximum shoulder IR and ER were measured at 90° of shoulder abduction with the elbow flexed to 90°. A bolster was placed under the distal humerus to maintain arm position in the plane of the scapula. The position of the scapula was controlled by palpating the coracoid process and monitoring for movement to determine the end range of glenohumeral motion. ²³ The inclinometer was placed along the ulnar shaft with the hand in a pronated position for IR and the inclinometer was placed along the ulnar border with the hand in the neutral position for ER. For all measurements, the inclinometer was placed proximal to the ulnar styloid. Horizontal adduction was measured in the supine position, with the arm brought across the body with the lateral border of the scapula stabilized by the examiners hand until resistance was met. The inclinometer was placed in the same position on the ulna and the measurement recorded. A single, board-certified orthopaedic surgeon performed ROM measurements of the elbow pain and asymptomatic cohort to reduce measurement variability, and another, board-eligible, orthopaedic surgeon recorded the measurements from the iPhone goniometer. The surgeon performing the measurements was blinded to the 2 testing groups. Two measurements were recorded for each ROM measure, and the average was then calculated and used for data analysis. GIRD and TAMD were calculated for all participants:

$$GIRD=\left(nonthrowingshoulderpassiveIRROM\right)–{\left(throwingshoulderpassiveIRROM\right)}^{\text{2}}$$

$$TAMD=\left(nonthrowingshoulderIR+ER\right)–\left(throwingshoulderIR+ER\right)$$

Test-retest reliability was performed on 7 individuals, prior to initiating the study, to determine intrarater reliability. The examiner reported excellent intrarater reliability for all ROM measures, with intraclass correlation coefficient (ICC) and minimal detectable change (MDC): IR—ICC(3, 2) = 0.94, MDC90 = 6.8°; ER—ICC(3, 2) = 0.91, MDC90 = 7.8°; GIRD—ICC(3, 2) = 0.95, MDC90 = 5.2°; TAMD—ICC(3, 2) = 0.996, MDC90 = 5.5°.

Statistical Analysis

Data were analyzed using Statistical Package for the Social Sciences (SPSS) software (Version 23; IBM Corp). Separate one-way analysis of variance was used to assess the shoulder ROM between pitchers with and without elbow pain. The level of significance level set a priori at P ≤ 0.05.

Results

There were no significant differences in age, height, and weight between pitchers with and without elbow pain. Descriptive shoulder ROM for GIRD, TAMD, and horizontal adduction between groups are reported in Table 2. Compared with the control group, the elbow pain group had a significantly greater GIRD (32.7° ± 13.4° vs 14.5° ± 9.7°; P < 0.05), and a significantly greater TAMD (28.3° ± 13.1° vs 6.7° ± 11.1°;P < 0.05).

Table 2.

Range of motion mean (SD) differences between the control and elbow pain groups

	Control Group	Elbow Pain Group
Dominant arm
Internal rotation a	59.6 (11.0)	34.0 (15.8)
External rotation a	122.1 (6.1)	104.5 (10.9)
Horizontal adduction a	40.4 (4.6)	45.2 (8.6)
Total arc of motion a	181.7 (8.8)	138.5 (22.2)
Nondominant arm
Internal rotation a	74.1 (9.4)	66.7 (9.6)
External rotation a	114.3 (6.3)	100.1 (10.5)
Horizontal adduction a	38.2 (6.8)	46.3 (7.9)
Total arc of motion a	188.4 (10.8)	166.8 (14.6)
Differences
Internal rotation a,b	−14.5 (9.7)	−32.7 (13.4)
Total arc of motion a,c	−6.7 (11.1)	−28.3 (13.1)

^aStatistically significant.

^bInternal rotation difference was calculated by the following formula: dominant arm internal rotation – nondominant arm internal rotation.

^cTotal arc of motion difference was calculated by the following formula: dominant arm total arc of motion – nondominant arm total arc of motion.

The prevalence of GIRD and TAMD was also higher in the elbow pain group compared with the control group (88% vs 33.3% and 96% vs 55.6%, respectively), as shown in Table 3.

Table 3.

Results

	Control Group (n = 18)	Elbow Pain Group (n = 25)
1. Internal rotation deficit ≥18°	33.3% (6 of 18)	88% (22 of 25)
2. Total range of motion deficit ≥5°	55.6% (10 of 18)	96% (24 of 25)

Discussion

GIRD and TAMD were significantly greater in the elbow pain group compared with the control group as hypothesized.

Garrison et al ⁷ reported no correlation between GIRD and elbow injuries in high school and collegiate baseball players but did associate TAMD with ulnar collateral ligament (UCL) tears. They found a TAMD of 6.67°, which is significantly less than the 28.3° TAMD found in youth pitchers presenting with elbow pain in this study. Garrison et al reported that 30% of their study participants were nonpitchers, which could affect shoulder ROM measurements since position players do not throw with the same frequency or velocity as pitchers. It is also unclear if differences in ROM measurements could account for this discrepancy: Werner et al ²² found excellent agreement between the smartphone clinometer and the gold standard goniometer-based measurement of shoulder ROM. Large standard deviations were observed in this study, and although significant differences were demonstrated, this would imply interpatient variability with unclear clinical significance.

Ostrander et al ¹⁵ reported on 108 male high school, college, and professional pitchers with UCL injuries compared with 108 asymptomatic, matched controls. Bilateral shoulder ROM measurements were obtained to assess for differences among the groups and pitching levels. They reported greater GIRD and TAMD in UCL-injured pitchers compared with the control group, which agrees with our findings. They did not find a variance of shoulder ROM deficits among different levels of pitchers but acknowledge that no other data in the literature exist to support or dispute this finding. Given the scarcity of data, more research is necessary in this area to provide definitive conclusions among any variation that exists between the levels of high school, college, and professional pitchers.

Meister et al ¹³ examined glenohumeral rotational ROM in adolescent baseball players aged 8 to 16 years and found glenohumeral forward flexion and total ROM decreased as age increased. The most dramatic decline in ROM was seen between 13- and 14-year-olds, in the year prior to the peak incidence of proximal humeral epiphysiolysis. They suggest that although retroversion changes should cause an increase in ER, there is also an overall loss of shoulder rotational motion occurring with skeletal maturity between the ages of 8 and 16 years. The decrease in total ROM may be attributed to loss of excessive tissue compliance from age-related changes in collagen as skeletal maturity approaches.

Nakamizo et al ¹⁴ reported a prevalence of GIRD in 40% (10 of 25) of asymptomatic youth pitchers aged approximately 11.5 years. These pitchers had developed GIRD prior to developing an increase in ER, which is contrary to what is seen in adult throwers. This finding is consistent with the theory that a tight posteroinferior capsule leads to GIRD.

Edelsonl ⁵ demonstrated that humeral retroversion reaches near adult parameters by age 8 years with a slowdown in the process between ages 8 and 16 years. The appearance of the radial groove was also noted at the end of this process, which appears between ages 12 and 16 years. Therefore, early recognition and intervention at a younger age appears pivotal to prevent GIRD and TAMD before final adult retroversion is achieved. The concern would be that the GIRD and TAMD could potentially be unresponsive to therapy and stretching exercises if the humeral head is in excessive retroversion. This demonstrates a limitation in the current study, and throughout the literature currently. GIRD is meant to imply a deficit in IR but should more appropriately indicate a difference in IR between dominant and nondominant arms because humeral retroversion is necessary to calculate a true GIRD. Hibberd et al ⁸ evaluated groups of healthy youth and adolescent baseball players using diagnostic ultrasound and found that GIRD and humeral retrotorsion increased with age, while retrotorsion-adjusted GIRD (true GIRD) and TAMD were unchanged. They concluded that an age-related increase in GIRD was primarily attributed to humeral retroversion, as opposed to soft tissue contracture. This study is limited by its use of only healthy baseball players, including both pitchers and position players, and is difficult to generalize to an injured population. Additional studies that assess humeral retrotorsion in an injured cohort may help better define their conclusion.

The nondominant arm had a greater total arc of motion in the control group, as well as an increased horizontal adduction in the group presenting with elbow pain. These findings may be related to an adaptation in youth pitchers due to the volume of throws and young age. As previously discussed, the majority of retroversion occurs by age 8 years, with the remainder developing until age 16 years. ⁵ Year-round baseball leads to a high volume of throwing with abduction–ER moments, which does not allow the proximal humerus to move from its normal 70° of retroversion seen in infants, to the normal adult parameters of 30°. ⁵

Given that these youth throwers fail to achieve normal adult parameters, which may account for the increased rates of shoulder and elbow injuries, these findings are likely multifactorial and a combination of bony adaptations and posterior capsular contracture.

Sabick et al ¹⁸ showed that shear stress in the late cocking phases was high enough to lead to deformation of the proximal humeral epiphyseal cartilage, thus leading to either humeral retroversion or proximal humeral epiphysiolysis. Furthermore, a significant positive correlation of adaptive morphological changes with humeral retrotorsion and glenoid retroversion has been reported in the dominant throwing shoulder of professional baseball players compared with their nonthrowing shoulder. This supports a possible mechanism leading to an increase in ER and potential decrease in IR and horizontal adduction. ²⁵ Reagan et al ¹⁶ evaluated a cohort of 54 asymptomatic college baseball players and reported a pattern of increased ER and decreased IR of the dominant throwing shoulder that significantly correlated with humeral retroversion, which was determined radiographically. They agreed that the loss of IR and gains in ER may be related to adaptive changes in proximal humeral anatomy. Additional research is needed to clarify the significance of a directed shoulder ROM physical therapy program given that some rotational ROM changes may be related to bony morphological adaptations and unlikely to respond to conventional, soft tissue–based therapy. It is possible that intervention prior to skeletal maturity may enable correction of soft tissue changes and minimize GIRD and TAMD.

One limitation of the current study is the small sample size, which may affect the overall power of the data. The design of our study included only pitchers with elbow pain and a matched, asymptomatic cohort, in an effort to minimize any confounding variables. We were also unable to account for individual variation in pitching frequency, which may also contribute to elbow pain. ¹¹

An additional limitation is the need to assess for humeral retroversion to determine a true GIRD. This calculation is most accurately determined with advanced diagnostic imaging, such as computed tomography or magnetic resonance imaging, and is not feasible for many clinicians to obtain on initial evaluation. Although ultrasound techniques have been described, this is still user dependent and may not be easily assessable for many clinicians to make these measurements. ²⁶

Conclusion

Compared with asymptomatic youth pitchers, those presenting with elbow pain have a statistically significant GIRD and TAMD. There is a high prevalence of GIRD and TAMD in youth pitchers presenting with elbow pain, with clinically significant differences compared with an asymptomatic, matched cohort. This serves to further support that shoulder ROM deficits may place throwers at an elevated risk for elbow injury. Further investigation should be aimed at true GIRD using a calculated humeral retroversion and determine for a response to directed physical therapy before skeletal maturity.