Increases in Ball Weight and Size Decrease Elbow Varus Torque During Baseball Pitching

Abstract

Background:

The rates of ulnar collateral ligament injury and surgery continue to rise in baseball. Increased ball velocity and elbow varus torque may correlate with the increased risk of injury.

Hypothesis:

Increased ball weight and/or size correlate with decreased elbow varus torque during pitching.

Study Design:

Controlled laboratory study.

Methods:

Motion capture data of 20 healthy professional and high-level collegiate baseball pitchers pitching fastballs with 4 types of baseballs were collected. The baseballs were 5 oz (standard weight) or 6 oz, with a circumference of 9 inches (standard size) or 5% larger. Five kinetic parameters, 25 kinematic parameters, and 7 ball movement parameters were calculated. Differences for each parameter were compared using 2-factor (ball weight × size) repeated-measures analysis of variance (P < .05).

Results:

As hypothesized, elbow varus torque decreased with increased ball weight and with increased ball circumference. Ball velocity, shoulder internal rotation velocity, elbow extension velocity, and shoulder kinetics also decreased with increased ball weight and/or increased ball circumference. Ball break decreased with increased ball weight, while ball location as it crossed home plate was also affected by ball weight and/or size. There were no clinically important differences in pitching kinematics with the different baseballs.

Conclusion:

Increasing the weight of baseballs from 5 oz to 6 oz and/or the circumference by 5% may reduce elbow varus torque. Future research in league play or simulated play is warranted.

Clinical Relevance:

As elbow varus torque has been staassociated with UCL injury risk, increasing ball weight and/or size may reduce the rate of injury.

The rate of ulnar collateral ligament (UCL) injuries in professional baseball continues to rise.^2,5,15 Because of this problem, Major League Baseball (MLB) surveyed more than 200 experts, including former professional pitchers, orthopaedic surgeons, athletic trainers, club officials, biomechanists, player agents, amateur baseball stakeholders, and other experts in pitcher development. This effort resulted in a new report that emphasized the roles of a continued increase in pitch velocity and a modern trend toward maximal effort pitching.^1,16 The report expressed a consensus that both are associated with high loads within the pitching elbow and the aforementioned high rate of UCL injury.^1,16 This theory is supported by biomechanical research, showing that (1) higher fastball velocity correlates with greater elbow varus torque, especially within-pitcher,^20,22 and that (2) higher elbow varus torque leads to a higher risk of UCL injury.^3,12 A prospective, longitudinal study of 305 professional pitchers demonstrated greater varus torque in pitchers with subsequent UCL surgery than in pitchers who did not end up sustaining a UCL injury. ¹² to Risk of UCL surgery increased 26% for every 10-Nm increase in elbow varus torque. Thus, reducing elbow varus torque may be a key to reducing the rates of UCL injuries.

The new MLB report suggests that professional baseball should consider rule changes to reduce the occurrence of UCL injury, with specific mention of shifting incentives and priorities for teams and pitchers toward health and longevity and away from the pursuit of maximal velocity and habitual pitching at maximal effort levels.^1,16 While the report mentioned the value of changes in amateur baseball, it stressed that amateur players generally model their own behavior around trends and incentives of the professional game, and thus initiatives at the professional level must lead the way.^1,16

Another possible area for MLB rule changes aimed at reducing elbow varus torque might be to adjust the fundamental properties of the baseball (eg, weight and size). A biomechanical study of 25 high school and college baseball pitchers showed that as ball weight increased from 5 oz (standard weight) to 6 to 7 oz, elbow varus torque significantly decreased. ⁹ Another study, with 26 college and professional pitchers, found that as ball weight increased from 5 oz to 6 to 7 oz, elbow varus torque decreased, but the decrease was not statistically significant. ¹⁸ A third set of researchers investigated the effects of reduced ball weight and size in youth pitchers, ¹⁷ citing how most other sports modify the size of their ball as players grow and progress through levels of competition. While only a few of their identified trends reached statistical significance, they suggested this may have been due to the lower kinetics of youth pitchers, compared with more physically mature pitchers competing at higher levels of the sport.

The mean size of top MLB pitchers has steadily increased since at least the 1950s, when players averaged 1.84 m (6 feet, 0 inches) in height and 83.9 kg (185 lbs) in weight, in comparison with 1.92 m (6 feet, 4 inches) and 100.3 kg (221 lbs) in the 2000s, ¹³ yet the regulation ball weight and size have remained unchanged during this time span. ²¹ Considering all of the above, the purpose of the current study was to investigate whether elbow varus torque is affected by ball weight or size in professional and high-level collegiate pitchers. It was hypothesized that elbow varus torque would be less when pitching 6-oz baseballs compared with when pitching 5-oz baseballs. It was also hypothesized that elbow varus torque would decrease if the ball size was increased to be 5% larger than the standard (9-inch circumference) size. Finally, it was hypothesized that there would also be significant differences related to both weight and size for other kinetic, kinematic, and ball movement parameters.

Methods

This study was approved by the Ascension St. Vincent's institutional review board. A total of 20 healthy collegiate baseball (Division I) and professional (AA minor league) baseball pitchers were recruited for the study. Inclusion criteria were being an active baseball pitcher for a collegiate or professional team, a self-reported mean fastball velocity of ≥87 mph during the most recent season, and self-reported as healthy and able to pitch at 100% effort at the time of testing. Exclusion criteria were elbow or shoulder surgery during the previous 12 months and being listed on an MLB team's 40-man roster at the time of recruitment. Data from a previous study ⁹ pitching 5-oz and 6-oz baseballs showed standard deviations of paired differences in elbow varus torque averaging 3.3 Nm. Assuming a similar value for this study, a sample size of 20 pitchers was determined to detect within-pitcher torque differences of ≥2.25 Nm between baseballs.

Four types of baseballs (Command Training Balls; Driveline Baseball) were used in this study, comprising all combinations of 2 ball weights and 2 ball sizes (Table 1). One ball type (A) represented a regulation baseball, with weight and size according to current MLB rules (5 oz; 9-inch circumference). The other ball types were 1 oz (28 g) heavier and/or 5% larger in diameter/circumference. Eighteen balls of each type were used during testing, with each ball type stored in its own container, separated from the other types. The mass of each individual ball was measured to the nearest gram, while the diameter was measured to the nearest hundredth of an inch (with the circumference subsequently calculated).

Table 1.

Four Types of Baseballs ^a

Ball	Manufacturer Description	Mass/Weight, g (oz)	Circumference, cm (inches)
A	5 oz (standard size)	138 ± 1 (4.86 ± 0.03)	22.7 ± 0.0 cm (8.9 ± 0.0)
B	5 oz, +5% size	144 ± 2 (5.10 ± 0.06)	23.8 ± 0.1 (9.4 ± 0.0)
C	6 oz (standard size)	168 ± 1 (5.94 ± 0.03)	22.7 ± 0.1 (8.9 ± 0.0)
D	6 oz, +5% size	167 ± 1g (5.91 ± 0.03)	23.8 ± 0.1 (9.4 ± 0.0)

^aData are presented as mean ± SD.

Participants were tested in pairs. Upon arrival at the biomechanics lab, each participant provided written informed consent. Pitchers were told that they would be throwing 4 short “innings” composed of 7 full-effort fastball pitches each, from an indoor mound, all aimed at the center of a strike zone located over a home plate at the regulation 60.5 feet (18.44 m) distance away from the pitching rubber. One pitcher was assigned to pitch the “top half” of each simulated inning (ie, pitch first) while the other was assigned to the “bottom half” (ie, pitch second). The ball type order for each pair of participants was predetermined with balanced randomization and the participants were blinded to the order and the ball type details.

A total of 47 reflective markers were then adhered to bony landmarks on each pitcher, using methods previously described.^6,8-11,22,23 Each pair of participants then warmed up playing catch together with a regulation baseball. The first participant then pitched as many warm-up pitches as he desired from the mound with the ball type that had been assigned to the first inning. When he indicated that he was ready, the participant pitched 7 fastballs for data capture. When the first participant was done with his half inning, the second participant threw as many warm-up pitches as desired with the first ball type and then his 7 fastballs for data capture. This procedure was then repeated for the 3 other innings with the 3 other ball types. Testing players in pairs created time between ball types for each pitcher to minimize any short-term effect of switching ball types. It was also similar to what pitchers experience in games, as 2 teams take turns pitching while the other team is batting.

Ball speed, location of the ball as it crossed home plate (“cross-plate location”), and ball break were recorded with a PITCHf/x system (Sports Media Technology). The signs/directions of horizontal location and break values were adjusted so that a positive value always denoted being more toward the pitcher's arm side. Motions of the reflective markers were captured with an automated digitizing motion analysis system (Motion Analysis Corporation). The automated system included 12 motion analysis cameras synchronized to measure the location of the reflective markers at 240 Hz. Raw data of the markers were filtered using a fourth-order Butterworth low-pass filter with a cutoff frequency of 13.4 Hz. Five kinetic parameters and 25 kinematic parameters were calculated for each trial, as previously described.^{6,8,10,11,22,23} In these previous studies, the ball was modeled as a 5-oz (142-g) point mass until it was released; in the current study, the ball was modeled as a 5-oz (142-g) or 6 oz (170-g) point mass until release, depending upon which ball was thrown.

Statistical Analysis

For each participant, the data from his first 2 pitches with each ball type were excluded from analysis to allow for any acclimation to the ball. Mean values of all parameters were then computed for the final 5 pitches for each ball. Differences for each parameter were compared among the 4 ball types using a 2-factor (ball weight × ball size) repeated-measures analysis of variance (ANOVA), with 2 levels for each factor. ANOVA results were first assessed to see whether there was a statistically significant between-factor interaction effect for that variable. If there was no interaction, main effects for each factor were assessed. For variables with a statistically significant interaction effect, simple main effects were assessed. The unadjusted threshold for statistical significance for all analyses was set at α = .05, with Bonferroni adjustments for multiple comparisons. All analyses were conducted in SPSS (Version 30.0; IBM Corp).

Results

The mean ± SD of age for the 20 pitchers was 21.5 ± 1.6 years. Height and mass for these participants were 1.90 ± 0.05 m and 95.3 ± 6.7 kg, respectively.

Mean values and standard error for kinetic parameters are shown in Table 2. Near the instant of maximal shoulder external rotation, elbow varus torque and shoulder internal rotation torque were significantly less with the increase in either ball weight or size. Shoulder horizontal adduction torque was also significantly less for both the (5-oz) 5% larger ball and the 6-oz (standard size) ball in comparison with the 5-oz (standard size) ball. Near the instant of ball release, shoulder proximal force was significantly less with the increase in ball weight.

Table 2.

Pitching Kinetics

	5 oz	5 oz, +5%	6 oz	6 oz, +5%	SE
Near the instant of maximal shoulder external rotation
Maximal elbow varus torque, Nm a b	99.5	97.3	98.2	95.6	3.1
Maximal shoulder internal rotation torque, Nm a b	100.6	98.3	99.0	96.5	3.0
Maximal shoulder horizontal adduction torque, Nm c d e	118.9 d e	113.4 e	115.4 d	115.0	4.2
Near the instant of ball release
Maximal elbow flexion torque, Nm	72.7	72.0	73.4	72.1	3.4
Maximal shoulder proximal force, N a	1227	1213	1208	1203	39

^aStatistically significant main effect (ball weight).

^bStatistically significant main effect (ball size).

^cStatistically significant interaction effect (weight × size).

^dStatistically significant simple main effect (weight) for the standard size balls.

^eStatistically significant simple main effect (size) for the 5-oz balls.

Mean values and standard error for kinematic parameters are shown in Table 3. Shoulder internal rotation velocity and elbow extension velocity were significantly less with the increase in either ball weight or size. While upper trunk angular velocity showed a similar trend, its decrease reached statistical significance for only the increase in ball size. Statistically significant differences were found for several angles, but the largest of these differences (the decrease in shoulder abduction at foot contact with increasing ball size) was only 1.1°.

Table 3.

Pitching Kinematics ^a

	5 oz	5 oz, +5%	6 oz	6 oz, +5%	SE
Instant of maximal knee height
Maximal knee height, % height	57.3	57.4	57.1	57.3	1.5
Instant of front foot contact
Elbow flexion, deg b	104.9	104.5	104.1	103.5	2.8
Shoulder abduction, deg b c	91.4	90.3	90.5	89.3	1.9
Shoulder horizontal abduction, deg	21.9	22.0	22.3	22.3	2.5
Shoulder ER, deg	67.6	68.3	68.2	68.5	5.0
Upper trunk tilt, deg b	7.9	7.9	7.5	7.2	1.7
Trunk axial rotation, deg b	48.4	47.8	47.4	47.2	1.6
Pelvic rotation, deg	34.4	35.1	34.4	35.5	3.0
Knee flexion, deg	44.1	44.4	44.8	44.4	1.8
Lead foot angle, deg	7.3	6.5	7.1	6.9	2.5
Lead foot position, cm b	17.4	16.9	17.8	18.2	2.7
Stride length, % height	82.5	82.1	81.7	81.9	1.3
Arm cocking phase
Maximal pelvic angular velocity, deg/s	565	561	563	560	15
Maximal upper trunk angular velocity, deg/s c	1173	1166	1170	1158	17
Instant of maximal shoulder ER
Elbow flexion, deg	115.2	115.6	115.2	115.0	2.6
Shoulder horizontal adduction, deg b	17.3	17.7	17.7	18.0	1.2
Maximal shoulder ER, deg c	166.8	166.2	166.8	166.1	1.7
Arm acceleration phase
Maximal elbow extension velocity, deg/s b c	2790	2752	2736	2717	81
Maximal shoulder IR velocity, deg/s b c	6718	6622	6546	6516	184
Instant of ball release
Arm slot from vertical, deg	54.7	54.3	54.4	55.2	2.6
Elbow flexion, deg b	27.2	27.4	28.0	27.5	1.1
Trunk forward tilt, deg	31.5	31.8	32.3	31.9	1.2
Trunk side tilt, deg	22.0	22.6	22.8	21.8	1.8
Shoulder abduction, deg	88.8	88.8	88.5	89.0	1.4
Knee flexion, deg	37.8	38.9	38.7	39.3	2.5

^aER, external rotation; IR, internal rotation.

^bStatistically significant main effect (ball weight).

^cStatistically significant main effect (ball size).

Mean values and standard error for ball movement parameters are shown in Table 4. Fastball pitch speed at both ball release and when crossing home plate were significantly less with the increase in either ball weight or size. Horizontal, vertical, and total ball break were significantly less with the heavier (6-oz) balls. An increase in ball weight and/or size led to a cross-plate horizontal location that was further to the throwing arm side, although only the ball size main effect reached statistical significance. An increase in ball weight and/or size led to a cross-plate vertical location that was significantly higher.

Table 4.

Ball Movement

	5 oz	5 oz, +5%	6 oz	6 oz, +5%	SE
Pitch speed at ball release, mph a b	85.1	83.9	82.3	81.3	0.6
Pitch speed crossing home plate, mph a b	75.0	74.2	73.9	73.1	0.5
Horizontal ball break, inches c de	11.3 d	11.1 e	8.2 d	8.8 e	0.7
Vertical ball break, inches a	12.2	11.9	9.5	9.2	0.8
Total ball break, inches a	17.4	17.1	13.4	13.4	0.6
Cross-plate horizontal location, inches b	2.0	4.0	3.4	7.5	1.5
Cross-plate vertical location, inches a b	25.1	32.9	36.1	43.2	1.8

^aStatistically significant main effect (ball weight).

^bStatistically significant main effect (ball size).

^cStatistically significant interaction effect (weight × size).

^dStatistically significant simple main effect (weight) for the standard size balls.

^eStatistically significant simple main effect (weight) for the 5% larger balls.

Discussion

As hypothesized, the increases in ball weight and/or size correlated with less elbow varus torque. Compared with the regulation baseball, the current study showed a 1.3% decrease in torque from increasing the weight to a 6-oz ball, a 2.2% decrease from increasing the size by 5%, and a 4.0% decrease from increasing both the weight and the size. In comparison, 2 previous studies each showed a 3% decrease in torque when increasing ball weight from 5 oz to 6 oz.^9,18 The decreased elbow varus torque in the current study was likely due to decreased angular acceleration, as suggested by the lower velocities (shoulder internal rotation, elbow extension, and ball velocities) with the increased ball weight and/or size. Additional shoulder kinetic variables also decreased with increased ball weight and/or ball size. While statistically significant differences were found for several body angles during pitching, their magnitudes were below the minimal clinically important difference of 2° previously set for baseball pitching ¹⁰ and likely have little practical importance.

Benjamin Franklin famously said, “An ounce of prevention is worth a pound of cure.” The current laboratory study found that an ounce of increased ball weight (28 g) and/or a 0.45-inch (1.1 cm) increase in ball circumference led to decreased elbow varus torque. If the decreased elbow varus torques reported in this laboratory study translate onto the field, such changes have potential to reduce the rate of UCL injuries. As shown in Figure 1, a prospective, longitudinal study reported that professional pitchers who sustained elbow injuries resulting in UCL surgery demonstrated significantly greater elbow varus torque (100.8 ± 18.1 Nm) than professional pitchers without such injury (94.3 ± 16.1 Nm). ¹² Results from the current study suggest increasing ball weight to 6 oz and/or ball size by 5% in circumference may decrease elbow varus torque from magnitudes similar to those shown by injured pitchers toward the magnitude of healthy pitchers (Figure 1).

Figure 1.

Mean values for elbow varus torque during pitching from prospective study ¹² and current study. Prospective study included pitchers who subsequently sustained an elbow injury resulting in ulnar collateral ligament surgery and pitchers who did not sustain such an injury. Current study shows elbow varus torque pitching regulation baseball (5 oz; 9-inch circumference) and balls of greater weight (6 oz) and/or size (+5%).

At first glance, the recommendation that increasing the weight of the baseball may reduce injury rates seems to contradict concerns of sports medicine professionals about increased injury risk from weighted ball training programs. For example, Reinold et al ¹⁹ trained adolescent baseball pitchers for 6 weeks with weighted baseballs and reported increased risk of elbow injury compared with their control group. In the Reinold et al study, the training program involved throwing balls ranging from 2 oz to 32 oz and the throwing drills were kneeling, rocking, and “run-and-gun” motions—not pitching. Thus, pitching with 6-oz, +5% baseballs is a significantly different activity than the wide range of ball weights and throwing exercises in weighted ball training programs.

Today’s professional baseball pitcher is bigger and stronger than ever before with greater pitch velocity as well. Greater ball velocity and pitching with maximal effort correlate with greater elbow varus torque and stress on the UCL.^20,22 Unfortunately, bigger body size does not correlate with a stronger ulnar collateral ligament. ⁴ Thus, today's baseball pitchers often produce high-magnitude elbow varus torque leading to UCL injury. ¹² A slightly larger and heavier baseball may be more appropriate for today's bigger and stronger pitchers. Interestingly, a study of players in one professional organization reported that pitchers with a history of UCL reconstruction had smaller hands than pitchers without history of UCL reconstruction. ¹⁴

While self-reported fastball velocity of ≥87 mph was required for inclusion, the mean fastball velocity during testing was 85.1 mph with the standard baseball. In our anecdotal experience, most pitchers throw about 3 to 7 mph slower in a laboratory than they do in a game. This is consistent with differences between bullpen testing velocity and self-reported velocity reported in high school pitchers. ⁷

However, there are several unknown factors about how changing the baseball may affect the rates of UCL injuries. First, the current study investigated the effects of ball weight and size for a homogeneous group of pitchers with similar body size and fastball velocity. The magnitude of change to elbow varus torque may vary somewhat for pitchers of different size and skill level. The magnitude of change to elbow varus torque may also vary for different types of pitches (eg, curveball, changeup, sweeper). Furthermore, this study measured the acute effects of varying ball weight and size. It is possible that pitching biomechanics may change over time when using heavier and/or larger baseballs. Pitchers may also make strategic alterations, such as change their pitch usage and pitch design to optimize “stuff ”. ¹⁶ Such adaptations may affect ball velocity, ball movement, kinematics, and elbow torque. The current study is a first step in exploring potential benefits of modifying the baseball. More biomechanical research is needed to investigate changes in pitching kinematics, joint kinetics, ball movement, and consistency over the course of a game and over the course of a season with heavier, larger baseballs. Advanced musculoskeletal modeling and simulation techniques could be utilized to investigate the load on the UCL itself, relative to the other musculoskeletal elements that can contribute torque about the varus/valgus axis. ²⁴ Eventually, clinical research should be conducted to investigate pitcher satisfaction, arm fatigue, and injury rates in league play with heavier, larger baseballs.

Decreased ball velocity and ball movement from adjusted ball weight and/or size could have a cascading effect on the outcomes and performance of baseball games. For example, decreased ball velocity and movement may result in more balls hit into play. Changes to ball weight and/or size may also affect the travel of batted balls (eg, exit velocity, hit distance) and the ease with which they can then be fielded. Ball weight, size, velocity, location, and movement may also affect the frequency and severity of batters hit by pitch and injured.

In addition to ball weight and size, adjustments to other physical characteristics of the baseball can be considered. The surface tackiness of the baseball, height of the seams, and ball hardness (coefficient of restitution) should be considered along with changing ball weight and size, and all these factors are worthy of future study. Potential modifications to a pitcher's grip in response to any of these changes may be of particular interest, as such adjustments could influence the distribution of varus torque among the UCL and other structures.^14,24

Conclusion

Increasing the weight of baseballs from 5 oz to 6 oz and the circumference by 5% may reduce elbow varus torque and the risk of elbow injury. Future research in league play or simulated play is warranted, including observations of the effects on hitting and fielding, as well as any long-term adaptation strategies developed by pitchers.