The institutional review board of Gunma University Hospital (Identification number 1003) approved this study. All methods were carried out in accordance with relevant guidelines and regulations. Written informed consent was obtained from the parents of the participants.
In this prospective study, we examined 128 high school male baseball pitchers who were 15 to 17 years of age. They participated in pre-season medical checkups, which were held in January and February 2018.
Pitchers who were able to complete a daily questionnaire about the presence of low back pain, which was collected every month, were considered for inclusion. Those who have had orthopedic surgeries before or already had shoulder, elbow or low back pain at pre-season medical checkups.
Pre-season medical checkups
As previously reported[17, 18], pre-season medical checkups were performed as baseline medical examinations. To avoid confirmation bias, participants’ hand dominance was not announced to examiners. The participants completed a questionnaire on their baseline characteristics, including height, their years of baseball experience, and hand dominance. The following physical parameters were assessed: (1) shoulder ROM and (2) shoulder muscle strength.
Shoulder ROM and strength measurements
The intra-rater and inter-rater reliability of digital protractors have been established in the literature[17]. According to previously reported methods[6–10, 16–18], a certified orthopedic surgeon assessed the passive shoulder ROM of 90° abducted external and internal rotation (ABER, ABIR) and horizontal adduction (HA), and the elbow ROM of flexion and extension were measured on both the dominant and non-dominant shoulders using a digital protractor with a bubble level indicator (iGaging, Los Angeles, CA, USA). The participant was placed in the supine position on the examination table, with the shoulder abducted to 90° and elbow flexed to 90º. A small rolled towel was placed under the elbow to keep the humerus in the right position. The scapula was stabilized posteriorly against the examination table by applying pressure to the coracoid process using the thenar eminence and thumb, and the humerus was passively rotated both ABER and ABIR until an end feel was obtained and the scapula began to move. The axis of the digital protractor was placed on the olecranon process of the elbow with the stationary arm aligned vertically, and the moving arm was aligned with the forearm. The total arc was calculated for the dominant shoulder by adding the ABER and ABIR.
When the shoulder HA ROM was measured, the participant was positioned in the supine position on the examination table with the scapula stabilized by the examiner’s pressure on the lateral border of the scapula with the thenar eminence. The test shoulder and elbow were positioned in 90° of both flexion and abduction. The examiner’s opposite hand then held the participant’s forearm, and the humerus was passively moved into HA until an end feel was obtained and the lateral border of the scapula began to move. The axis of the digital protractor was placed at the estimated center of the glenohumeral joint with the stationary arm perpendicular to the horizontal plane, and the moving arm was aligned with the humerus. All shoulder ROM measurements were performed by two examiners with one examiner providing stabilization force to maintain the shoulder position while the other obtained the ROM measurement.
The intra-rater and inter-rater reliability of the shoulder strength measurements by hand-held dynamometers have been established in a previous study[17]. In accordance with previous studies[2, 17, 18], using a PowerTrack II Commander hand-held dynamometer (J-Tech Medical, Salt Lake City, UT, USA), a certified orthopedic surgeon measured the strength of the supraspinatus in the seated position (SS), prone external rotation (PER) and prone internal rotation (PIR) in both the dominant and non-dominant shoulders. When the SS strength was measured, the participant sat on the examination table with his back against the wall. The humerus was abducted to 90° and then horizontally adducted to 45° with the forearm neutral. The examiner placed the dynamometer 5 cm proximal to the proximal wrist extension crease, and the participant raised his arm perpendicular to the floor with maximum effort. The PER and PIR strength were measured in the prone position with the shoulder abducted to 90º and the elbow flexed to 90º. The examiner stabilized the humerus and set the arm in 0º of rotation, and then the participant rotated his arm externally and internally with maximum effort against the dynamometer. When the PIR strength was measured, the dynamometer was placed 5 cm proximal to the proximal wrist flexion crease, and when the PER strength was measured, the dynamometer was placed at the dorsal side of the forearm, opposite to the proximal wrist flexion crease.
Low back injury
In this study, “low back injury” was defined as any condition that resulted in the pitcher being considered disabled for eight days or more[17]. In the statistical analyses, other injuries caused by other mechanisms, such as being hit by a ball, colliding with other players, or suffering trauma from falls, were excluded. To avoid recall bias, participants were instructed to complete a self-recorded questionnaire everyday regarding the presence of low back pain, limitations to pitching caused by low back pain and the presence of other injuries.
Statistical analyses
Statistical analyses were performed using the SAS 9.4 software program (SAS Institute Inc., Cary, NC, USA). All tests were two-sided with a P = 0.05 significance level. Depending on the presence of low back injuries, the participants were divided into injured and non-injured groups. Categorical data were reported as the frequency (%), and group differences were evaluated using the chi-square test. Continuous data were reported as median with inter quarter range (IQR), and group differences were evaluated using the Mann-Whitney U-test. After adjusting for significant variables identified in univariate analyses, the logistic regression analysis was performed to calculate odds ratios (ORs) and 95% confidence intervals (CIs) to identify the risk factors for low back pain.
To determine the sample size for this study, a prior statistical power analysis for a logistic regression analysis was performed. This analysis indicated that a total of 70 participants would be needed, depending on a statistical power of 80% at an α level of 0.05 (assumptive incidence rate = 20%, OR ratio = 2.5)[20].