Within the sport of baseball, performance metrics have been examined in attempt to mitigate injury susceptibility in youth pitchers, including number of pitches thrown, innings pitched, and rest days [22]. Unlike baseball, softball does not have an official workload recommendation; however, some teams anecdotally attempt to quantify and limit pitch count and workload. Since both sports require proximal to distal sequencing within an open kinetic chain, and injury prevention measures have not thoroughly been investigated within the sport of softball, further investigation into not only workload but also anterior shoulder forces over an extended workload is necessary.

With the increased awareness of injury susceptibility in softball, there has been a shift in focus to examining associations between the prevalence of pain and pitching mechanics [23, 24]. It has been found that those pitching with upper extremity pain have altered pitching mechanics compared to those pitching without pain [23]. Similarly, the association of an increased stride length and a posteriorly shifted center of mass (COM) with those experiencing upper extremity pain has been established in collegiate pitchers [23, 24]. However, to the authors’ knowledge, there are no known studies examining pain and pitching mechanics in youth softball athletes. With the increased awareness of pain and fatigue in youth softball pitchers [1–3, 25–27], as well as the lack of established sanctions of pitch count regulations, consecutive innings and/or games pitched, or required rest days, greater investigation into youth pitching and the acute biomechanical changes that occur following a pitching performance is needed. Previously, the examination of the LHBT in youth revealed significant increases in longitudinal and transverse thickness following an acute bout of pitching. These dimensional increases were hypothesized to be an acute inflammatory response [28] which emphasizes the stress that the LHBT endures over the duration of a game. While research has not yet determined the specific cause of LHBT inflammation during a pitching bout, there is reason to believe that mechanics, especially those surrounding the trunk and upper extremity, may impact the degree of change in LHBT measures. Specifically, certain pitch mechanics may elicit a greater inflammatory response than others, which may cause more undue harm and shoulder pain. With pitchers already displaying high rates of shoulder pain, it is pertinent to understand how mechanics may exacerbate the stress at the shoulder, which is currently believed to cause shoulder pain.

Therefore, it was the purpose of this study to examine the association of acute changes in the LHBT, with pitching kinematics and kinetics in youth softball pitchers following an acute bout of pitching. Specifically, we investigated the association of LHBT changes (longitudinal thickness, transverse thickness, transverse width, and transverse area) with trunk kinematics (flexion, lateral flexion, and rotation), COM (in relation to base of support), stride length (height normalized), and shoulder kinetics (abduction/adduction moment; internal/external moment, and flexion/extension moment) prior to and immediately following a simulated game protocol. It was hypothesized that LHBT thickness (longitudinal and transverse), width (transverse), and area (transverse) would increase following the simulated game when compared to pre-simulated game measurements. It was also hypothesized that the aforementioned increase would be associated with altered pitching kinematics (trunk, COM, and stride length) and increased shoulder kinetics (abduction/adduction moment; internal/external moment, and flexion/extension moment) between the first and last inning of a simulated game.

Materials and Methods

Twenty-three female youth softball pitchers (12.17 ± 1.50 yrs.; 160.32 ± 9.41 cm; 60.40 ± 15.97 kg) volunteered to participate. Inclusion criterion required the participants to be actively competing on a team’s roster as a pitcher. Additionally, they had to be surgery and injury-free for the past six months. Injury was defined as being diagnosed by a physician or athletic trainer resulting in time loss from practice or competition. The Institutional Review Board of Auburn University approved all testing protocols and informed written consent was obtained from each participant prior to participation.[29] Participants were instructed on the LHBT imaging and simulated game protocol. A NextGen LOGIQe Ultrasound (GE Healthcare, Milwaukee, WI, USA) using a 4–12 MHz linear array transducer in B-mode was used for LHBT image collection. Acquisition parameters were constant for all participants and included a 10 MHz frequency, 3.5 cm depth, and one transmit focus point [28].

Each participant had dominant arm LHBT ultrasound imaging prior to and immediately following the simulated game protocol. Ultrasound data were collected using previously established methods [28, 30]. The participant was seated with elbow flexed, forearm supinated, and wrist in neutral resting on the contralateral knee. The investigator placed the probe on the anterior aspect of the shoulder on the LHBT, perpendicular to the humerus. All measurements were taken at the widest qualitatively identifiable aspect of the LHBT. Transverse and longitudinal views were recorded as single images. The transverse view was used to locate the bicipital groove through the identification of the greater and lesser tuberosity. Once the bicipital groove was determined, an image of the LHBT was recorded. The probe was then turned 90°, while remaining at the same level of the transverse view, to record the longitudinal view [28, 30]. The investigator marked a small dot on the skin at the lateral border of the transducer to identify the bicipital groove. This procedure was repeated three times and the mean of the three measurements was used for analysis.

Two investigators performed all ultrasound measurements. Intra-rater reliability was determined during a pilot study of 5 participants. Measurements of transverse width, transverse thickness, and longitudinal thickness were performed using the diameter function on the ultrasound machine. Transverse area was measured by tracing the tendon with the ultrasound trackball. The investigators reported excellent intra-rater reliability using the techniques previously described, with an ICC(3,k) of 0.90–0.98 for all measurements. Standard error of measurement (SEM = pooled standard deviation × √1-ICC) and the minimal detectable change at the 90% confidence interval (MDC₉₀ = SEM × √2 × 1.65) for transverse width, transverse thickness, and longitudinal thickness were also calculated from the pilot data. For the data to indicate a significant change not related to measurement error, observed differences had to exceed the MDC. Longitudinal thickness MDC₉₀ was 0.024 cm, transverse width MDC₉₀ was 0.014 cm, transverse thickness MDC₉₀ was 0.029 cm, and transverse area MDC₉₀ was 0.019.

Kinematic data were collected at 240 Hz using an electromagnetic tracking system (trackSTAR™, Ascension Technologies, Inc., Burlington, VT, USA) synced with Biomechanics Analysis Software (The MotionMonitor, Innovative Sports Training, Chicago, IL, USA). Eleven electromagnetic sensors were attached to the participants using previously established methodologies [31]. A linked segment model was developed using the digitized joint centers for ankle, knee, T12-L1, and C7-T1 by the digitized medial and lateral aspect of each joint, then calculating the midpoint between those two points [32–35]. Hip and shoulder joint centers were estimated by means of a previously established rotation method [36, 37]. The world axis was represented with the positive Y-axis in the vertical direction; anterior to the Y-axis and in the direction of movement was the positive X-axis; and orthogonal and to the right of X and Y was the positive Z-axis. Raw data regarding sensor positioning and orientation were transferred to a locally based coordinate system. Euler angle sequences consistent with the International Society of Biomechanics standards and joint conventions were used to define position and orientation of the body segments [34, 35]. The trunk was modeled relative to the world axis, using the Euler sequence of ZX′Y″, while the YX′Y″ sequence was used for shoulder motion relative to the trunk. All raw data were independently filtered along each global axis using a 4^th-order Butterworth filter with a cut off frequency of 13.4 Hz [32, 33, 38].

Once all kinematic sensors were secured, participants were given an unlimited time to perform their own specified pre-competition warm-up (average warm-up time was 7 minutes). Participants were instructed to throw their desired pitches (fastball and change-up) based on randomly provided game situations. An expert with seven years of youth, high school, and collegiate coaching experience developed the game situation protocol [39, 40]. The investigator provided verbal feedback of hitter counts (balls and strikes), simulated at-bat outcomes (base hit, base on balls, hit-by-pitch, ground-outs, and fly-outs), and simulated runners. Participants were required to produce three outs an inning as per the standard rules of softball. After three outs were recorded, a rest period was provided to simulate the second half of the inning. In attempt to mimic typical offensive innings in youth softball, rest periods were randomly altered in length from 7 to 14 minutes. Pitch count was limited to the participant’s average pitch count during competition. Average pitch count for the simulated game was 61 ± 7 pitches. It should be noted that though the participants threw a simulated game, no fatigue measures were assessed. Three fastballs in the first inning and three fastballs in the last inning were recorded, averaged, and used for analysis. The pitching motion was broken into five different events; Start of pitching motion, top of pitch foot contact (FC), ball release (BR), and follow-through. All kinematic and kinetic variables were analyzed at each of the five pitching events.