Utilizing the Pronated Forearm Technique for Measuring Glenohumeral External Rotation in Baseball Players

Abstract

Background:

Alterations in glenohumeral internal rotation (GIR), glenohumeral external rotation (GER), and total range of motion (TROM) have been linked with increased injury risk. GER capacity has been measured routinely with the forearm in neutral rotation (GERN), but a recent study reported GERN was greater than GER with the forearm in pronation (GERP) in Minor League pitchers. This work has not yet been replicated or extended to other groups.

Hypothesis:

GERP would be significantly less than GERN in Independent League baseball pitchers, and there would be no difference in GERP or GERN measurements between this new group and the previous group of Minor League pitchers.

Study Design:

Cross-sectional study.

Level of Evidence:

Level 3.

Methods:

Goniometric measurements were recorded for bilateral GIR, GERN, and GERP, and resulting TROM for 37 Independent League baseball pitchers. These data were compared with the previous study. All motions were compared individually between groups, between throwing and nonthrowing arm, and both within and between techniques (forearm neutral or pronated).

Results:

GERP was significantly less than GERN for both arms within each group tested (P < 0.01). Independent League pitchers had greater between arm differences for GIR (-16.9° vs -6.9°), GERN (+15.1° vs -0.6°), and GERP (+13.1° vs -5.9°) compared with Minor League pitchers. TROM for the Independent League pitchers was not statistically different for either measurement technique, while TROM for the throwing arm of the Minor League pitchers was statistically reduced with varying effect sizes (d = 0.35-0.99) compared with the nonthrowing arm (P < 0.01).

Conclusion:

This study confirmed earlier findings that the pronated forearm resulted in decreased GER capacity, illustrating the adaptive response to throwing and the need to evaluate for this variable.

Clinical Relevance:

GERP should be evaluated in all groups of pitchers, but there may be variations within tested groups.

Mobility deficits can serve as possible sources of shoulder pain due to negative changes in the viscoelastic properties of tendons and/or alterations in muscle or tendon structure due to degeneration or repetitive load. Alterations in glenohumeral internal rotation (GIR), glenohumeral external rotation (GER), and total range of motion (TROM) have been linked with increased injury risk in the shoulder and elbow in various overhead athletes providing clinical justification for measuring rotational range of motion objectively in these groups.^{1,5,9,11,17,24,30,31,37, 39} Several anatomic components can be related to mobility deficits, including humeral retrotorsion, muscle thixotropy and stiffness, and capsular thickening.^{3,10,17,20,23,25 -28,33,35,36} However, the muscular components are most often considered clinically likely due to the immediate changes that occur when muscle is exposed to loads from activity or rehabilitation,^14,15,18,29 but it is important to distinguish between normal variants and pathological mobility deficits.

Manske et al ¹⁸ noted that anatomic glenohumeral internal rotation deficit (GIRD) is an entity that is normal in overhead athletes and is characterized by internal rotation loss <18° to 20° with symmetrical total arc of motion bilaterally. Conversely, pathologic GIRD is defined as a loss of GIR >18° to 20° with a corresponding loss of total arc of motion >5° when compared bilaterally. Wilk et al ³⁹ have determined that insufficient GER (defined as not having more GER on the throwing arm compared with the nonthrowing arm) is associated with increased injury risk, while Camp et al ¹ reported decreased GER can be a greater predictor of injury than GIR deficits. These clinical phenomena were further supported by a meta-analysis that concluded that professional baseball players were at 2 times risk of injury if the throwing side to nonthrowing side difference in GER was not at least >5° to 8°. ²⁴ Because of these findings, more attention has been placed on evaluating and optimizing GER in injury risk modification programs.

Anecdotal observation of arm and forearm position in throwing highlighted that the forearm was in a position of pronation in several important parts on the entire motion, and that including forearm position during passive GER measurements could be a key variable that may influence the objective measurement being used to determine injury risk. Following up on those observations, Kibler et al ¹³ reported significantly higher GER values when measuring GER with the forearm in neutral (GERN) versus measuring with the forearm pronated (GERP). It was concluded that the GERP measurement should be considered as part of measurement screenings because it aligns with biomechanical occurrences during overhead throwing and better estimates the external rotation capacity. However, that study was the first on the topic and was exclusive to pitchers and position players in Minor League baseball. Therefore, the purpose of this study was to replicate the original methodology focusing on a separate cohort and examiner. In addition, the study aimed to compare values obtained in the current cohort with values from the original study. The hypotheses were as follows: (1) GERP would be significantly less than GERN and (2) there would be no difference between the cohorts in GERP or GERN measurements.

Methods

Pitchers from 2 independent baseball teams were assessed before the beginning of the Spring 2022 season. Age, height, weight, throwing arm, position, and current injury were recorded. This study was a cross-sectional analysis of shoulder rotation measures before the beginning of a single competitive season. The study was reviewed and approved by an Institutional Review Board (Approval No. LCO-2022-001).

Shoulder measurements were taken during preparticipation physical examinations, before any baseball or conditioning activities occurred, to minimize the possibility of alteration of range of motion from those activities. GIR, GERN, and GERP were measured bilaterally with a standard bubble goniometer. Each pitcher was placed in a supine position on a flat level surface. A second examiner was positioned behind the pitcher to properly stabilize the scapula during testing by applying a posteriorly directed force to the coracoid and scapula to ensure that scapular movement did not occur.^14,41 The humerus was supported on the surface with the elbow placed at 90°, the arm on a bolster in the plane of the scapula, and the forearm in a position of neutral rotation. The following landmarks were identified before placing the goniometer: the fulcrum was set at the olecranon process of the elbow, the stationary arm perpendicular to the table as documented by the bubble on the goniometer, and the moving arm in line with the styloid process of the ulna. Each subject was then advised to relax, while the humerus was moved passively into internal rotation. Rotation was taken to “natural tightness,” ie, the point of passive motion where no more glenohumeral motion would occur unless overpressure (pressure applied that would possibly move the joint past the achieved end point) was applied or the scapula would move (Figure 1). The humerus was then moved into an externally rotated position to the “natural tightness” position (GERN) (Figure 2). The arm was not forced at the end of the motion, with no applied pressure, but was guided by the examiner and allowed to move to its endpoint. The examiner was permitted to rest his fingers of the hand holding the moving arm of the goniometer on the forearm of the subject to help align and steady the goniometer. The measure was recorded when humeral motion ceased, and the goniometer was stable according to the bubble. The arm was then returned to the starting position. The final measurement followed the same procedures except that the forearm was placed in a full pronated position and held in this position before passively guiding and allowing the arm to move to the endpoint of external rotation (GERP) (Figure 3). The procedures were performed bilaterally to obtain measurements from both the throwing and nonthrowing shoulder. Elbow pronation/supination range of motion measurements were not obtained in this study.

Figure 1.

Internal rotation measurement.

Figure 2.

External rotation measurement with neutral forearm.

Figure 3.

External rotation measurement with pronated forearm.

Statistical Analysis

Descriptive/summary statistics were performed for all demographic variables and were reported as means and standard deviations for continuous variables such as age, height, and weight, while counts and percentages were used for throwing arm. The distribution of data for each functional variable was assessed for normality using the Shapiro-Wilk test.

The baseline data from the current study cohort were compared with historical baseline data made using the same measurement techniques on a cohort of pitchers from Minor League baseball. ¹³ All motions were compared individually between groups, between throwing and nonthrowing arm, and both within and between techniques (forearm neutral or forearm pronated) using appropriate pairwise procedures (paired or independent t tests). TROM for each shoulder was calculated as the addition of GIR and GERN (TROMN) as well as GIR and GERP (TROMP). Alpha was set a priori as P ≤ 0.05.

Pairwise Cohen’s d coefficients were calculated to determine the relative effect size for individual variables that were statistically significant. ⁴ The effect size is often used to determine whether mean differences are large enough in magnitude to be considered clinically meaningful, and Cohen defined effect sizes as small, d ≤ 0.4, medium, d = 0.41-0.79, or large, d ≥ 0.8.^4,7 All analyses were performed with STATA 17.0/SE (STATA Corp).

The test/retest reliability of the single examiner was determined for all shoulder measurements. Intraclass correlation coefficients were calculated in a 2-way random effects model and absolute agreement. The standard error of measurement and minimal detectable change at the 90% and 95% confidence level were also determined.

Results

All intraclass correlation coefficients were >0.75, indicating excellent test/retest reliability (Table 1).^2,19 Post hoc power analysis comparing the differences between GERP and GERN values in the current cohort revealed that the study was adequately powered (beta, 0.98) with effect sizes ≥0.74.

Table 1.

Reliability analysis for range of motion measurements

	ICC (95% CI)	SEM	MDC90	MDC95
GIR	0.84 (0.32, 0.96)	4.2	9.7	11.5
GER, neutral	0.95 (0.80, 0.99)	2.7	6.2	7.4
GER, pronated	0.83 (0.37, 0.96)	4.1	9.5	11.3

GER, glenohumeral external rotation; GIR, glenohumeral internal rotation; ICC, intraclass correlation coefficient; SEM, standard error of measurement; MDC₉₀, minimal detectable change at the 90% confidence level; MDC₉₅, minimal detectable change at the 95% confidence level.

The baseline motion measurements for 37 Independent League pitchers (age, 28.4 ± 3.7 years; height, 188.2 ± 6.5 cm; weight, 95.9 ± 8.9 kg, 78% right-handed) was compared with baseline measurements of 32 Minor League baseball pitchers (age, 24.1 ± 2.0 years; height, 188.9 ± 4.5 cm; weight, 90.0 ± 9.8 kg, 69% right-handed) reported previously in the literature. ¹³ The Independent League pitchers were significantly older (P < 0.01) and had greater mass (P = 0.02).

All motions for both arms of the Minor League pitchers were statistically reduced with large effect sizes (d ≥ 0.79) compared with the Independent League pitchers (P < 0.01) except for GIR of the throwing arm (Table 2). Independent league pitchers had greater between-arm differences for GIR (-16.9° vs -6.9°), GERN (+15.1° vs -0.6°), and GERP (+13.1° vs -5.9°) compared with Minor League pitchers (Table 3). However, TROM for the Independent League pitchers was not statistically different for either measurement technique whereas TROM for the throwing arm of the Minor League pitchers was statistically reduced compared with the nonthrowing arm (-7.5° for TROMN and -12.7° for TROMP, P < 0.01). GERP measures were significantly less than GERN for both arms within each group (P < 0.01): Minor League pitchers throwing arm, -9.7° and nonthrowing arm, -4.4°; Independent League pitchers throwing arm, -6.1° and nonthrowing arm, -4.1°. However, the effect sizes were variable, ranging from low to large effects (d = 0.35-0.99).

Table 2.

Baseline motion comparisons between cohorts ^a

	Minor League Pitchers (n = 32)	Independent League Pitchers (n = 37)	Difference	P Value	Effect Size
GIR
Throwing	29.2 (9.2)	28.2 (10.6)	-1.0	0.69	0.10
Nonthrowing	36.1 (8.7)	45.1 (13.2)	-9.0	<0.01	0.79
GER, neutral
Throwing	79.7 (9.3)	112.5 (10.2)	-32.8	<0.01	3.35
Nonthrowing	80.3 (13.2)	97.4 (8.8)	-17.1	<0.01	1.55
GER, pronated
Throwing	70.0 (10.3)	106.4 (10.7)	-36.4	<0.01	3.46
Nonthrowing	75.9 (12.0)	93.3 (10.3)	-17.4	<0.01	1.56
TROM, neutral
Throwing	108.9 (8.5)	140.7 (12.7)	-31.8	<0.01	2.90
Nonthrowing	116.4 (10.6)	142.5 (14.1)	-26.1	<0.01	2.07
TROM, pronated
Throwing	99.3 (9.5)	134.6 (13.2)	-35.3	<0.01	3.03
Nonthrowing	112.0 (9.7)	138.4 (15.3)	-26.4	<0.01	2.03

Values are reported in degrees. GER, glenohumeral external rotation; GIR, glenohumeral internal rotation; TROM, total range of motion.

^aReported as mean (SD).

Table 3.

Motion comparisons between arms and measurement techniques ^a

	Throwing	Nonthrowing	Difference	P Value
Minor League pitchers (n = 32)
GIR	29.2 (9.2)	36.1 (8.7)	-6.9	<0.01*
GER, neutral	79.7 (9.3) b	80.3 (13.2) b	-0.6	0.71
GER, pronated	70.0 (10.3)	75.9 (12.0)	-5.9	<0.01*
Effect size	0.99	0.35
TROM, neutral	108.9 (8.5) b	116.4 (10.6) b	-7.5	<0.01*
TROM, pronated	99.3 (9.5)	112.0 (9.7)	-12.7	<0.01*
Effect size	1.07	0.43
Independent League pitchers (n = 37)
GIR rotation	28.2 (10.6)	45.1 (13.2)	-16.9	<0.01*
GER, neutral	112.5 (10.2) b	97.4 (8.8) b	+15.1	<0.01*
GER, pronated	106.4 (10.7)	93.3 (10.3)	+13.1	<0.01*
Effect size	0.58	0.43
TROM, neutral	140.7 (12.7) b	142.5 (14.1) b	-1.8	0.37
TROM, pronated	134.6 (13.2)	138.4 (15.3)	-3.8	0.56
Effect size	0.47	0.28

Values are reported in degrees. GER, glenohumeral external rotation; GIR, glenohumeral internal rotation; TROM, total range of motion.

^aReported as mean (SD).

^bNeutral position significantly greater compared with pronated position (P < 0.01).

^*Significant difference (P ≤ 0.05).

Discussion

There is evidence to accept 1 of our 2 hypotheses following replication of a previous study’s methodology for obtaining rotational range of motion in overhead athletes. There is evidence to accept the hypothesis that GERP would be significantly less than GERN, with the differences being similar to the original study. This suggests that the forearm pronation component creates similar physiological changes across populations, validating the technique’s ability to provide different information about GER capacity compared with the traditional neutral forearm position component. However, the second hypothesis that there would be no difference between the cohorts in GERP or GERN measurements was rejected. The Independent League pitchers had significantly greater between-arm differences for GIR, GERN, and GERP. This suggests that cohort/population differences should be accounted for when interpreting rotational range of motion of the shoulder.

Methodological replication is invaluable to science. Replication helps confirm or refute reported findings, which in turn bolsters evidence to verify or dismiss possible clinical phenomena.^6,22 It could be argued the current study was not true replication because it sampled a similar (baseball pitchers) but not identical cohort (age and mass were statistically greater compared with the historical cohort). However, the time of year the measurements were obtained, identical measurement techniques as described by the previous authors, and that the examiner from the current study (who was different from the previous study) had demonstrated excellent test/retest reliability with the measurement techniques support that replication occurred. These factors help place the findings in context but the possible reasons for why the findings occurred should be discussed.

The identified differences both within and between the cohorts could have 2 clinical implications. First, the Independent League players had the theorized GIR loss balanced by GER gain whereas Minor League players did not.^18,40 This could be due to variations in throwing volume/frequency, off-season training, and possible anatomic differences. For example, the Minor League pitchers were relatively new to the professional level, suggesting they probably had recent high school and/or college level playing exposure as well as structured training occurring in the professional organization. Most Independent League pitchers had previous professional experience with Minor League and/or Major League Baseball, with their most recent time in those systems ranging from months to years. At the time of testing, several of the Independent League pitchers were or had been recently employed in nonbaseball/nonathletic vocations and anecdotally mentioned occasional training. The lack or reduced amount of recent repetitive overhead throwing/training in the Independent Pitcher cohort could account for the balanced lower GIR and higher GER seen in that group. Although training load or frequency of either cohort were unknown, workload has been identified as a factor for injury risk and may be a component to account for in future investigations.^16,21,32,34 Similarly, the GIR and GER values reported by Kibler et al ¹³ for the Minor League pitchers that were compared with the values in the current study appear to be low compared with most reported literature.^{8,29,37 -39,41} It is possible that variations in measurement technique between the studies could contribute to this discrepancy. The technique employed in this study of taking the arm to “natural tightness” was initially described by Kibler et al ¹⁴ as part of their prospective longitudinal study that examined range of motion changes in professional pitchers over 72 hours after pitching during spring training. The lower motion values could be related to this technique, considering that other works with reported higher motion values did not describe withholding overpressure to the arm during measurements of motion.

Second, the differences between arms in each cohort also resulted in differences among reported injury risk thresholds.^24,31,37,38 Neither cohort exceeded the suggested anatomic internal rotation deficit of 18° to 20° between arms. ¹⁸ However, the Minor League pitchers were within the 5° to 8° injury risk threshold for side-to-side external rotation differences with both measurement techniques (-0.6° for GERN and -5.9° for GERP). These deficits in turn resulted in an exceeding of the injury risk threshold of 5° for TROM (-7.5° for TROMN and -12.7° for TROMP). Conversely, the balance between GIR loss and GER gain among the Independent League pitchers resulted in TROM differences between arms that were minimal and not statistically different. These findings indicate that (1) although side-to-side GIR differences receive much attention, side-to-side GER values should not be overlooked ¹ ; and (2) overhead athlete injury risk may not be universal across populations, even in similar sports.

Similar to the original study, the current investigation was cross sectional in its design, so injury occurrence was not monitored prospectively. Although it is well documented that range of motion alterations are a risk factor for injury in overhead athletes, ³⁴ it is unknown whether the risk of injury is for time-loss or nontime-loss injury. Epidemiological data have noted that approximately 60% to 85% of all injuries seen in high school and college baseball qualify as nontime-loss injuries, with only 18% to 23% of those injuries occurring at the shoulder. ¹² The sports medicine staff informed the research team that only 2 time-loss injuries occurred all season in the current cohort (1 biceps tendonitis and 1 vascular thoracic outlet syndrome). However, it is unknown in the current or original study whether nontime-loss injury occurred; therefore, any future attempts to replicate this methodology should consider whether relationships exist between GERN/GERP values and injury occurrence/risk accounting for both time-loss and nontime-loss injuries.

Limitations

There are limitations to this study to acknowledge. First, because of the intent to replicate the methodology from an existing study, no modifications were employed to address the limitations noted by the previous authors in their work. These limitations included not measuring forearm motion (pronation/supination) and not determining whether biceps tightness or potential reasons for biceps tightness were present. It is possible these anatomic factors could influence the results regarding the consistently decreased GER values with the pronated forearm position compared with the neutral forearm position. Second, the original study reported range of motion values of pitchers and position players as individual groups and collectively, whereas the current work focused only on comparing pitchers from the previous study with pitchers in the current cohort. This was intentional because, at the time of testing, the team rosters were not complete as several position players had not yet arrived at the team location due to travel, immigration, and other logistical reasons. Thus, position players were not included in the current study. Finally, although a different cohort was sampled that was statistically older and heavier, the cohort was similar in that all pitchers had previous professional baseball experience. It is possible the results could vary if replicated in younger populations, different sexes, and different overhead sports.

Conclusion

After replicating the methodology from a previous study for obtaining shoulder rotational range of motion using varying forearm positions, the clinical relevance of the findings is that forearm position does affect the measurement of GER and GER capacity. These findings support measuring GER with the forearm both in neutral and pronation, and TROM should be calculated from these same forearm positions. In addition, cohort/population differences can influence the results, suggesting that clinicians consider these factors when interpreting range of motion values. These recommendations are classified as Strength of Recommendation Taxonomy B based on the lower quality diagnostic case-control design.