Biomechanical coordination and variability alters following repetitive movement fatigue in overhead athletes with painful shoulder

Abstract

Most sports and leisure activities involve repetitive movements in the upper limb, which are typically linked to pain and discomfort in the neck and shoulder area. Movement variability is generally expressed by changes in movement parameters from one movement to another and is a time-dependent feature of repetitive activities. The purpose of this study was to examine the effect of repeated movement-induced fatigue on biomechanical coordination and variability in athletes with and without chronic shoulder pain (CSP). In this controlled laboratory study, 24 overhead athletes in two groups of athletes with (N = 12) and without (N = 12) CSP were recruited. Biomechanical and Electromyographical data were recorded while the athletes were asked to perform repeating reaching tasks (RRT). Kinematic data was recorded every 30 s of the minutes of the repetitive pointing task (RPT). The kinematic and electromyography data were recorded at the first 30 s of “Fatigue-Terminal” (FT) and the last 30 s of “No-Fatigue” (NF) in the repetitive pointing task (RPT). Raw data was analyzed by using MATLAB code to extract the relevant coordination and movement variability data. Different fatigue conditions led to significant kinematic changes during the repetitive pointing task. In the CSP group, trunk lateral flexion decreased after fatigue, while it increased in the CON group (p < 0.013). Trunk rotation and shoulder elevation angles were smaller before fatigue than after in both groups (p < 0.001). Variability in joint angles, including trunk lateral flexion and elbow flexion, increased after fatigue, indicating less stability in movement patterns (p < 0.001). The coordination between trunk movement and shoulder elevation was altered post-fatigue, with significant changes in EMG variability for muscles like the Lower Trapezius and Long Head of the Biceps (p < 0.001). Results of our study indicate that both groups were able to accomplish fatigue, but they employed different movement strategies. The CSP group primarily focused on controlling the shoulder joint, while the CON group utilized both the shoulder and elbow joints in their strategy. This difference suggests that individuals in the CSP group who experienced chronic pain may have developed a strategy to minimize pain and fatigue during the task. Specifically, the CSP group’s approach appeared to involve adjustments in movement patterns to manage the task despite fatigue. This adaptation contrasts with the CON group’s more complex movement strategy, which involved greater variability and adjustments in both the shoulder and elbow joints.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-85226-5.

Introduction

Repetitive movements in the upper limb are common in many sports and leisure activities¹, and are often associated with pain and discomfort in the neck and shoulder area². Several time-dependent physical factors are linked to Chronic Shoulder Pain (CSP), including prolonged postures and repetitive activities³. These repetitive movements, sometimes referred to as repetitive movement disorders (RMDs), can cause damage and pain to nerves or joints through chronic overuse injuries^1,3. Recent epidemiological studies on repetitive movements and chronic injuries indicate that 75% of all subtle RMDs occur in the upper limb⁴. However, the existing literature predominantly focuses on the lumbopelvic region⁵, leaving a significant gap in understanding the impact of RMDs on the neck and shoulder⁶. Performance typically declines with repeated muscle activity, which is attributed to fatigue rather than merely the repetitiveness of the movement⁷.

The scapula, clavicle, and humerus must coordinate to move the shoulder girdle, which can be described using two-dimensional (D2) and three-dimensional (D3) measurement methods^8,9. The ability to organize proper relationships between brain signals and joints during movements is known as coordination, and it is a key mechanism for producing smooth and precise movement patterns¹⁰. Thus, during repetitive tasks, inter-segment coordination may serve as a useful indicator of motor control systems¹¹. Several studies have shown that information about the timing of movements and the locations of two nearby segments, combined with kinematic analysis, can provide valuable insights into the kinematics of limb motions and the complexity of human motor behaviors^12,13. Movement variability, which refers to the natural fluctuations in movement patterns, can also offer crucial insights into motor control. Research has demonstrated that both chronic pain and repetitive movements can impact the variability and coordination of movement¹⁴.

A decrease in a muscle or muscle group’s ability to generate force following an activity is known as muscle fatigue¹⁵. Muscle fatigue resulting from repetitive tasks may affect motor coordination¹⁶. Individuals with coordination problems may be more prone to injuries due to poor biomechanics¹⁷. Conversely, increased muscular force variability related to force levels from muscle fatigue can lead to greater kinematic and kinetic variability¹⁸. Movement variability, defined as changes in movement parameters from one movement to another, is a time-dependent feature of repetitive activities¹⁹and is considered a potential determinant in the development of RMDs³. Additionally, reducing co-activation caused by muscle fatigue may increase kinematic variability²⁰. The idea is that higher co-contraction-related muscle activation results in increased signal-dependent noise and, therefore, higher variability is observed²¹. These studies suggest that a reduction in movement speed and an increase in baseline EMG activity might be associated with efforts to maintain motor efficiency and reduce muscle soreness rather than causing dissociation of coordination. However, findings indicate that greater co-contraction results in higher torque and EMG variability and lowers the end-point kinematic variability in targeted reaching movements¹⁶. Moreover, it seems that people may adjust their muscle activation patterns or biomechanical coordination when performing multi-joint dynamic tasks²², which can mitigate changes in overall kinematic coordination and variability²³.

It seems that a decrease in movement speed²⁴and an increase in background EMG activity related to muscle stability²⁵are associated with the presence of chronic pain²⁶. Individuals with persistent neck and shoulder pain are more likely to exhibit dissociation of joint movements from the kinematic chain during continuous, repetitive activities²⁷. Therefore, they may experience increased movement variability due to changes in their kinematic patterns. Although limited research has explored the link between pain and fatigue in inter-segmental coordination and variability among athletes with painfull shoulder, it seems that variability during daily activities, such as upper extremity repetitive movements, is an essential part of the motor control strategy. Although knowing the changes associated with chronic shoulder pain can be effective for designing protocols to prevent sports injuries or to rehabilitate injuries, however, no study has yet investigated the effect of fatigue on this issue. Therefore, the purpose of this study is to address the research gap by examining the impact of repeated movement-induced fatigue on biomechanical coordination and variability in athletes with and without chronic shoulder pain (CSP).

Methods

Participants

Overhead male athletes aged between 20 and 35 who have at least three years of history of regular exercise (at least 3 sessions per week) were recruited for this study. Twenty-four male overhead athletes participated in this study, divided into two groups: those with chronic shoulder pain (N = 12) in their dominant hand and those without (N= 12). The results of a study that examined the effect of fatigue on the electromyographic function of the lower trapezius muscle were used to estimate the required sample size in this study [30]. Using G*Power ver 3.1 software and considering an alpha of 0.05 and a beta of 0.20, the minimum number of subjects required for this study was 22 (11 in each group). One additional subject was included in each group to reduce the possible effect of possible dropouts. The inclusion criteria were elite male overhead athletes aged 20 to 35 who had been on the national team for over four years. Exclusion criteria included¹: a history of upper limb musculotendinous rupture or bony fractures², any exercise performed within 24 h before the test session³, a history of metabolic diseases such as diabetes⁴, use of sedative medications or medications that could impair nervous system function during the last week, and⁵ evident upper limb or trunk malalignment disorders based on the New York criteria (score 1 out of 5). An informed consent form was provided to each participant before testing began. The Allameh Tabataba’i University Ethics Committee in Tehran, Iran, approved the study on 25 January 2023 (IR.ATU.REC.1401.084).

Experimental protocol

Participants performed a repetitive reaching task (RRT) for as long as possible. The dominant hand was used for the task. In the RRT, subjects stood upright and reached back and forth between two cylindrical touch-sensitive targets positioned at 30% and 100% of arm length, shoulder height, and in front of the subject’s midline (Fig. 1). The targets were 6 cm in length, with a radius of 0.5 cm, and had a response time of 130 ms. An elliptically shaped mesh barrier was placed under the elbow joint’s functional range of motion to ensure that arm movement remained in the horizontal plane at shoulder height during the task. Participants used their index fingers to gently contact each target while keeping their elbows above and away from the mesh barrier. The RRT involved continuous movement from one target to the next. Subjects received audio feedback upon touching each target and used a metronome to maintain a rhythm of one movement per second (1 Hz).

Fig. 1.

Repetitive Pointing Task equipment.

During the RRT, participants recorded 30 s of electromyography (EMG) data at a sampling rate of 1200 Hz after rating their perception of task difficulty (Borg CR-10 scale) and CSP on a 0–10 Numeric Rating Scale (NRS) during the last 30 s of each minute of reaching. The RRT was conducted continuously until the participants could no longer maintain the 1-Hz rhythm or rated their discomfort as 8 or above on the NRS or Borg CR-10 scales. These stoppage criteria were not disclosed to the participants²⁸.

To help participants maintain a 1-Hz movement rhythm, a MATLAB program introduced a 1-second delay in optical stimuli and provided auditory feedback when the cylindrical touch-sensitive targets were illuminated. Data were collected every five seconds during 30-second intervals of the RPT.

Data acquisition

During the RPT, a VICON 8-camera motion analysis system was employed to monitor movement kinematics. Kinematics were recorded using 35 markers (diameter 12.7 mm) placed on bony landmarks based on three-dimensional coordinates. The markers were attached to the skin using hypoallergenic elastic tape, avoiding large muscle bellies to minimize marker movement during the task. Markers were first placed and recorded in a static position to define the kinematic model, which divided the body into 15 segments: head (head to C7; 5 markers), arms (shoulder to elbow; 3 markers on each upper arm), forearms (elbow to the wrist; 4 markers on each forearm), hands (distal part to the wrist; 3 markers on each hand), trunk (C7-T10; 5 markers), pelvis (5 markers), thighs (hip to knee; 3 markers on each thigh), legs (knee to ankle; 3 markers on each leg), and ankles (3 markers on each leg)²⁹ (Fig. 2). To calculate the values ​​of the variables required in this study, kinematic and electromyographic data were recorded throughout the RRT test. Then, the first 30 s of the test were analyzed as “No-Fatigue” (NF) and the last 30 s before reaching peak fatigue were analyzed as “Fatigue-Terminal” (FT) data.

Fig. 2.

Placement of the motion capture markers and EMG electrodes.

After skin preparation, disposable Ag/AgCl electrodes were applied in a bipolar configuration parallel to muscle fibers with a 3 cm center-to-center spacing, according to standard recommendations³⁰. parallel to the muscle fibers. For the lower trapezius muscle (TrapL), the electrodes were positioned approximately 25 mm from the midline of the body, near the inferior-posterior region of the shoulder, along a line drawn from the T8 vertebra to the medial angle of the scapula. These electrodes were placed close to the midline-medial area of the shoulder and scapula. For the anterior deltoid (DeltA), the electrode was placed 2 cm below the outer third of the clavicle. The electrode for the biceps long head (BicLong) was positioned in the middle of the arm’s upper section, at the muscle’s center. For the medial head of the triceps long head (TriLong), the electrode was placed 2 cm medial, vertical, and dorsal to the arm at the midpoint between the acromion and the olecranon^29,30. The Lower Trapezius and Anterior Deltoid were selected for their role in shoulder height activities³¹. At the same time, the Biceps and Triceps were included for their agonistic roles in shoulder and elbow joints during the RPT. Maximal Voluntary Isometric Contraction (MVIC) tests were conducted 30 min before the RPT to normalize the data as a percentage of MVIC. Subjects performed a maximal isometric contraction for the lower trapezius muscle by pulling the shoulder blades downward and inward. For the anterior deltoid muscle, subjects were seated or standing with the shoulder in a neutral or slightly flexed position. For the biceps long head, subjects positioned their arm in a flexed and supinated position and performed a maximal isometric contraction by flexing the elbow while keeping the shoulder neutral. For the triceps long head, subjects were positioned with the shoulder flexed and the elbow extended. They performed a maximal isometric contraction by extending the elbow against resistance, ensuring the shoulder remained neutral. Each muscle was contracted isometrically for 5 s, with three repetitions and a 60-second rest between contractions. EMG data were recorded from the reaching side, with efforts made to minimize cross-talk between muscles.

A TeleMyo 900 EMG measurement device (Noraxon USA Inc., Arizona, USA) was used for recording data, and a 16-bit A/D card transformed the 1200 Hz sampled data digitally for later analysis. Participants were instructed to continue touching and alternating between forward and backward phases. The forward phase began when the fingers touched the proximal target, and the distal target marked the end of the phase. The reverse phase was defined by the reverse of these events.

Data analysis

All EMG signals were first filtered using a fourth-order zero-phase Butterworth band-pass filter with 20–500 Hz cut-off frequencies. Following filtering, the overall mean was removed to center the data. The signals underwent full-wave rectification to further process the signals and eliminate heart rate artifacts, converting all values to their absolute magnitudes. The rectified signals were then integrated using the Average Rectified Value (ARV) method with a moving window of 0.025 s (25 milliseconds) and a step rate of 1 sample per window. To normalize the raw EMG data, each signal was expressed as a percentage of the MVIC of the corresponding muscle. Finally, for each marked forward movement, the root mean square (RMS) amplitude of the EMG from four muscles was calculated using MATLAB software (MathWorks Inc., Natick, MA, USA).

Kinematic data were recorded every 30 seconds during the RPT. For analysis, the first 30 seconds of “Fatigue-Terminal” (FT) and the last 30 seconds of “No-Fatigue” (NF) were used. Coordination between adjacent upper limb segments was analyzed using MATLAB software. Joint angles were computed based on Euler angles, dividing the body into trunk-pelvis, shoulder, and elbow segments³². The shoulder angles were defined as follows: the first rotation (Y’) corresponds to the plane of elevation; the second rotation (X) represents elevation; and the third rotation (Y’’) denotes axial rotation (internal/external rotation). The kinematics for the trunk and pelvis were also analyzed, with CRP (Continuous Relative Phase) used to assess coordination between shoulder and elbow and shoulder and trunk³³. CRP calculations involved determining the phase angle difference between proximal and distal joints, with 0° indicating in-phase movement and 180° indicating antiphase³⁴. Joint angles and CRP values were time-normalized to 101 points per cycle. The mean and standard deviation (SD) for these points were computed, excluding the first and last 5 cycles to avoid boundary issues. Regarding EMG variability, the mean EMG signals and SDs for the 5 cycles of each participant for all four muscles were calculated for the entire cycle and each phase.

Statistical analysis

Data were analyzed using IBM^®SPSS version 24.0 (IBM Corp, Armonk, NY, USA). The Shapiro-Wilk test assessed data normality. Independent Samples T-tests compared demographic data between groups, and One-way repeated measures ANOVA tested the results. The significance level was set at 0.05 for all analyses. Partial Eta Squared (ES) was used to assess clinical significance, with ES values between 0.5 and 0.8 indicating a substantial clinical difference and values over 0.8 indicating a large clinical difference³⁵.

Results

The demographic data for the study participants are summarized in Table 1. Notably, there were no significant differences in age, height, body mass, and BMI between the two groups.

Table 1.

Participant demographics.

Variable	CON (N = 12)	CSP (N = 12)	P-value
	Mean ± SD	Mean ± SD
Age (years)	21.66 ± 2.01	23.25 ± 2.76	0.123
Height(cm)	183.33 ± 12.82	182.58 ± 6.90	0.860
Body Mass (kg)	78.50 ± 12.70	72.66 ± 8.74	0.204
BMI (kg.m2)	23.33 ± 2.78	21.77 ± 2.08	0.134
VAS (mm)	N/A	2.83 ± 0.83	-

NF, FT stands for No-Fatigue and Fatigue-Terminal condition respectively. CON, Control Group; CSP, Chronic Shoulder Pain Group; BMI, Body Mass Index; VAS, Visual Analogue Scale of Current Pain. ** statistically significant differences between the CON and CSP groups, P<0.05.

One-way repeated measures ANOVA was run to analyze the data. The results demonstrate that different fatigue conditions (non-fatigue and terminal fatigue) led to distinct kinematic effects during the performance of the repetitive pointing task (refer to Table 2; Fig. 3). Specifically, fatigue-induced changes were evident in mean angles at all three joints. Notably, the mean angle for trunk lateral flexion in the CON group was greater after fatigue than before fatigue. Similarly, the CSP group exhibited a smaller mean angle after fatigue than before fatigue in the effect of time × group (p < 0.013). Additionally, the mean rotation angle for the trunk was greater before fatigue than after fatigue in the CON group, as indicated by the effect of time (p < 0.006) and the effect of time × group (p < 0.006). These changes imply that the trunk leaned less but rotated more toward the non-moving arm’s side after fatigue.

Table 2.

Joint angles result under all conditions (NF vs. FT) and groups (CON vs. CSP).

Variables		CON (N = 12)		CSP (N = 12)		Time Effect				Time × Group
		Mean ± SD		Mean ± SD
		NF	FT	NF	FT	F	P-value	ES	Power	F	P-value	ES	Power
Trunk (degree)	Lateral Flexion (left(-)/ right(+))	2.34 ± 0.99	3.38 ± 1.33	4.19 ± 1.50	3.36 ± 1.27	0.103	0.751	0.005	0.061	7.271	0.013**	0.248	0.732
	Rotation (left(-)/ right(+))	4.00 ± 0.87	2.66 ± 1.07	3.58 ± 1.44	3.59 ± 1.62	9.074	0.006**	0.292	0.821	9.349	0.006**	0.298	0.832
	Flexion(+)/ Extension(-)	5.16 ± 2.67	6.58 ± 3.45	4.10 ± 2.21	3.84 ± 2.57	1.654	0.212	0.070	0.233	3.538	0.073	0.139	0.436
Shoulder (degree)	Flexion(+)/ Extension(-)	41.56 ± 8.89	64.42 ± 8.77	35.87 ± 6.36	61.50 ± 9.76	111.314	0.001**	0.835	1.000	0.363	0.553	0.016	0.089
	Abduction(+)/ Adduction (-)	63.88 ± 13.05	43.00 ± 3.41	65.21 ± 9.29	58.92 ± 13.54	19.344	0.001**	0.468	0.987	5.579	0.027**	0.202	0.617
	Internal(-)/ External(+) Rotation	17.14 ± 3.21	14.19 ± 1.87	19.35 ± 2.66	18.68 ± 2.68	5.376	0.030**	0.196	0.601	2.149	0.157	0.089	0.289
Elbow (degree)	Flexion(+)/ Extension(-)	68.64 ± 3.60	77.16 ± 9.51	66.30 ± 9.06	69.14 ± 9.87	9.265	0.006**	0.296	0.829	2.315	0.142	0.095	0.307
	Pronation (-)/ Supination (+)	3.14 ± 1.12	2.69 ± 1.07	3.36 ± 1.50	3.09 ± 2.01	0.728	0.403	0.032	0.129	0.048	0.829	0.002	0.055
	Internal(-)/ External(+) Rotation	54.45 ± 13.28	66.41 ± 6.88	58.59 ± 6.81	72.09 ± 9.04	18.415	0.001**	0.456	0.984	0.068	0.796	0.003	0.057

NF, FT stands for No-Fatigue and Fatigue-Terminal condition respectively. ES, Effect Size; CON, Control Group; CSP, Chronic Shoulder Pain Group. **indicates a statistically significant interaction effect for the ANOVA (p<0.05).

Fig. 3.

Joint angles between all conditions (NF vs. TF) and Groups (CON vs. CSP). NF and TF represent non-fatigued RPT and terminal-fatigued, respectively. In the context of angle definitions based on research by Gates et al. (2016) and We et al. (2005), adjustments are made to the signs of various joint angles. Specifically, positive values are assigned to trunk lateral flexion, trunk rotation, and trunk flexion when bending toward the non-reaching side, rotating toward the reaching side, and bending forward, respectively. Similarly, positive shoulder elevation and shoulder rotation angles correspond to humerus horizontal flexion forward, humerus elevation, and humerus external rotation. Additionally, positive elbow flexion, elbow abduction, and forearm rotation indicate forearm flexion, forearm abduction, and pronation. * indicates a Time Effect for the ANOVA (p < 0.05), and ** indicates a Time × Group interaction for the ANOVA (p < 0.05).

Turning our attention to the shoulder, we observed that the mean plane of elevation angle in the CON group before fatigue was smaller than after fatigue. Similarly, the CSP group also had a smaller mean angle before fatigue compared to after fatigue in the effect of time (p < 0.001). Shoulder elevation angle in both groups before fatigue was greater than after fatigue in the effect of time (p < 0.001) and the effect of time × group (p < 0.027). Additionally, the shoulder rotation angle in both groups before fatigue was greater than after fatigue in the effect of time (p < 0.030). These findings suggest that the humerus was less forward and less elevated after fatigue than before.

Furthermore, the mean elbow flexion and rotation angles in both groups were also greater after fatigue than before fatigue in the effect of time (p < 0.006 and p < 0.001), indicating that, on average, the elbow was more flexed after fatigue.

The examination of joint angular variabilities (as detailed in Table 3; Fig. 4) investigated the impact of fatigue on the trunk and upper limb movement patterns. Trunk lateral flexion variability in the CON group increased after fatigue compared to the pre-fatigue condition. In contrast, the CSP group exhibited smaller variability after fatigue than before fatigue in the effect of time×group (p < 0.001). Additionally, trunk rotation variability in both groups was greater after fatigue than before fatigue (effect of time, p < 0.010). These increased trunk angular variabilities imply a less stable trunk movement pattern following shoulder fatigue.

Table 3.

Joint angular variabilities result under all conditions (NF vs. FT) and groups (CON vs. CSP).

Variables		CON (N = 12)		CSP (N = 12)		Time Effect				Time × Group
		Mean ± SD		Mean ± SD
		NF	FT	NF	FT	F	P-value	ES	Power	F	P-value	ES	Power
Trunk (degree)	Lateral Flexion (left(-)/ right(+))	0.96 ± 0.27	1.17 ± 0.23	1.57 ± 0.23	1.31 ± 0.07	0.215	0.647	0.010	0.073	16.443	0.001**	0.428	0.972
	Rotation (left(-)/ right(+))	0.83 ± 0.12	0.89 ± 0.26	0.74 ± 0.15	0.85 ± 0.12	7.833	0.010**	0.263	0.763	0.561	0.462	0.025	0.111
	Flexion(+)/ Extension(-)	0.84 ± 0.29	0.69 ± 0.31	0.81 ± 0.26	0.84 ± 0.31	0.484	0.494	0.022	0.102	1.185	0.288	0.051	0.181
Shoulder (degree)	Flexion(+)/ Extension(-)	7.51 ± 1.75	8.38 ± 2.02	6.93 ± 0.67	8.65 ± 0.73	22.082	0.001**	0.501	0.994	2.428	0.133	0.099	0.320
	Abduction(+)/ Adduction (-)	3.56 ± 1.05	4.04 ± 0.90	8.55 ± 2.05	5.39 ± 2.82	5.363	0.030**	0.196	0.600	9.910	0.005**	0.311	0.853
	Internal(-)/ External(+) Rotation	2.63 ± 1.06	2.38 ± 0.33	2.65 ± 0.87	3.16 ± 0.28	0.537	0.537	0.018	0.092	3.412	0.078	0.134	0.423
Elbow (degree)	Flexion(+)/ Extension(-)	23.65 ± 1.38	32.26 ± 4.12	28.61 ± 1.13	31.05 ± 1.42	49.637	0.001**	0.693	1.000	15.461	0.001**	0.413	0.964
	Pronation (-)/ Supination (+)	21.08 ± 0.10	21.07 ± 0.30	21.03 ± 0.36	21.14 ± 0.45	0.716	0.406	0.032	0.128	1.189	0.287	0.051	0.181
	Internal(-)/ External(+) Rotation	33.64 ± 4.00	25.97 ± 1.18	27.81 ± 3.29	29.32 ± 1.15	10.818	0.003**	0.330	0.882	24.108	0.001**	0.523	0.997

NF, FT stands for No-Fatigue and Fatigue-Terminal condition respectively. ES, Effect Size; CON, Control Group; CSP, Chronic Shoulder Pain Group. **indicates a statistically significant interaction effect for the ANOVA (p<0.05).

Fig. 4.

Joint angular variabilities between all conditions (NF vs. FT) and Groups (CON vs. CSP). * indicates a Time Effect for the ANOVA (p < 0.05), and ** indicates a Time × Group interaction for the ANOVA (p < 0.05).

After fatigue, the variability of the shoulder plane of elevation increased in both groups (effect of time, p < 0.001). This indicates altered shoulder movement patterns due to fatigue. Specifically, in the CON group, shoulder elevation angular variability after fatigue was greater than before fatigue. In contrast, the CSP group exhibited a decrease after fatigue compared to before fatigue in the effect of time (p < 0.030) and the effect of time × group (p < 0.005).

Elbow flexion variability in both groups increased after fatigue (effect of time, p < 0.001; effect of time×group, p < 0.001). Notably, in the CON group, elbow rotation variability before fatigue was greater than after fatigue. However, in the CSP group, it was greater after fatigue than before fatigue in the effect of time (p < 0.003) and the effect of time × group (p < 0.001). These findings highlight how fatigue at the trunk level can influence movement patterns of the distal joints (elbow) within the multi-joint linkage.

The investigation of mean CRPs and CRP variabilities (as detailed in Table 4; Fig. 5), investigated how fatigue affects joint kinematics. Specifically, we examined the relationship between trunk flexion/extension and shoulder plane of elevation in both groups before and after fatigue.

Table 4.

Mean CRPs and CRP Variabilities under all conditions (NF vs. FT) and groups (CON vs. CSP).

Variables		CON (N = 12)		CSP (N = 12)		Time Effect				Time × Group
		Mean ± SD		Mean ± SD
		NF	FT	NF	FT	F	P-value	ES	Power	F	P-value	ES	Power
ShPE-ElFl (degree)	Mean CRP	5.23 ± 0.68	4.90 ± 0.45	5.49 ± 0.92	5.22 ± 0.80	4.942	0.037**	0.183	0.566	0.055	0.816	0.003	0.056
	Variability	0.90 ± 0.08	0.92 ± 0.09	1.12 ± 0.21	1.01 ± 0.28	3.446	0.077	0.135	0.427	5.043	0.035**	0.186	0.574
TrFl-ShPE (degree)	Mean CRP	5.33 ± 0.88	4.42 ± 0.63	5.33 ± 0.71	5.19 ± 0.60	7.745	0.011**	0.260	0.758	4.230	0.052	0.161	0.503
	Variability	0.91 ± 0.17	0.87 ± 0.04	0.91 ± 0.11	0.94 ± 0.09	0.058	0.812	0.003	0.056	1.055	0.315	0.046	0.166
TrRo-ShPE (degree)	Mean CRP	5.30 ± 0.98	4.99 ± 0.53	5.26 ± 0.59	5.00 ± 1.07	2.200	0.152	0.091	0.295	0.018	0.894	0.001	0.052
	Variability	0.93 ± 0.21	0.97 ± 0.08	1.19 ± 0.16	1.04 ± 0.30	1.860	0.186	0.078	0.257	5.647	0.027**	0.204	0.622
TrLaFl-ShEl (degree)	Mean CRP	5.15 ± 0.54	4.81 ± 0.70	5.20 ± 0.60	4.50 ± 0.59	8.631	0.008**	0.282	0.802	1.097	0.306	0.047	0.171
	Variability	0.87 ± 0.06	0.90 ± 0.13	1.04 ± 0.16	0.88 ± 0.16	2.826	0.107	0.114	0.362	7.361	0.013**	0.251	0.737

NF, FT stands for No-Fatigue and Fatigue-Terminal condition respectively. ShPE-ElFl, TrFl-ShPE, TrRo-ShPE, and TrLaFl-ShEl indicate CRP between trunk flexion/extension and shoulder plane of elevation, CRP between trunk flexion and shoulder plane of elevation, CRP between trunk rotation and shoulder plane of elevation, CRP between trunk lateral flexion and shoulder elevation, respectively. ES, Effect Size; CON, Control Group; CSP, Chronic Shoulder Pain Group. **indicates a statistically significant interaction effect for the ANOVA (p<0.05).

Fig. 5.

Comparison of Mean CRPs and CRP Variabilities under All Conditions (NF vs. FT) and Groups (CON vs. CSP) in shoulder, elbow, and trunk motion before and after reaching fatigue between two groups.

Before fatigue, the mean CRP and CRP variability between trunk flexion/extension and shoulder plane of elevation exceeded the after-fatigue values (p < 0.037 and p < 0.011, respectively). This suggests altered coordination between trunk movement and shoulder elevation due to fatigue. Interestingly, in the CON group after fatigue, the CRP variability between trunk rotation and shoulder plane of elevation was greater than before fatigue. However, in the CSP group, this variability was greater before fatigue than after fatigue (effect of time × group, p < 0.027).

The mean CRP between trunk lateral flexion and shoulder elevation in both groups before fatigue exceeded the post-fatigue values (effect of time, p < 0.008). This indicates fatigue-induced changes in the alignment between lateral trunk movement and shoulder elevation. Furthermore, CRP variability in the CON group after fatigue was greater than before fatigue. In contrast, in the CSP group, this variability was greater before fatigue than after fatigue (effect of time × group, p < 0.013).

Regarding the coefficient of variability of the EMG signals (as presented in Table 5; Fig. 6), we observed that the coefficient of variability of the Lower Trapezius muscle in both groups increased after fatigue compared to the pre-fatigue condition. Specifically, this increase was significant in terms of both the effect of time (p < 0.001) and the effect of time × group (p < 0.028). Furthermore, we also found that the variability of the Biceps Long Head muscle in both groups was greater after fatigue than before fatigue, as indicated by the effect of time (p < 0.002).

Table 5.

Coefficient of variability (CV) of the EMG signal result under all conditions (NF vs. FT) and groups (CON vs. CSP).

Variables	CON (N = 12)		CSP (N = 12)		Time Effect				Time × Group
	Mean ± SD		Mean ± SD
	NF	FT	NF	FT	F	P-value	ES	Power	F	P-value	ES	Power
DeltA (% CV)	23.00 ± 3.64	22.00 ± 2.49	23.66 ± 6.62	25.00 ± 5.92	0.176	0.676	0.008	0.069	0.488	0.492	0.022	0.103
TrapL (% CV)	6.83 ± 3.78	12.83 ± 3.73	4.91 ± 2.77	7.00 ± 3.56	23.635	0.001**	0.518	0.996	5.549	0.028**	0.201	0.615
BicLong (% CV)	5.41 ± 1.72	7.00 ± 3.01	3.66 ± 1.15	5.83 ± 2.20	12/493	0/002**	0/362	0/922	0/302	0/588	0/014	0/082
TriLong (% CV)	4.00 ± 1.53	3.58 ± 2.23	3.41 ± 1.31	2.66 ± 1.07	2/015	0/170	0/084	0/274	0/689	0/698	0/007	0/067

NF, FT stands for No-Fatigue and Fatigue-Terminal condition, respectively. DeltA, Deltoid Anterior; TrapL, Lower Trapezius; BicLong, Biceps Long Head; TriLong, Triceps Long Head, respectively. ES, Effect Size; CON, Control Group; CSP, Chronic Shoulder Pain Group. **indicates a statistically significant interaction effect for the ANOVA (p<0.05).

Fig. 6.

Comparison of Coefficient of variability of the EMG signal changes in shoulder motion before and after reaching fatigue between two groups.

Discussion

The current study’s findings indicate that different fatigue conditions (non-fatigue and terminal fatigue) led to significant kinematic changes during the performance of the repetitive pointing task. The CSP group exhibited smaller variability after fatigue than before, highlighting a shift in movement strategy. Moreover, altered coordination between trunk movement and shoulder elevation was recorded after fatigue. Additionally, the variability in shoulder muscle activation in both groups differed significantly after fatigue.

It is important to note that inter-segment coordination data following fatigue—brought on by repetitive motions in athletes with and without pain—showed that the shoulder’s chronicity did not exceed 0.05 both before and after the fatigue limit. This suggests that chronic pain’s impact on shoulder biomechanics does not drastically alter the coordination metrics between segments like elbow-wrist and shoulder-elbow in a statistically significant manner over time. However, a statistically significant difference was found in the coordination between segments such as elbow-wrist and shoulder-elbow, driven by time and the interaction between time and group.

The unique characteristics of athletes necessitate a detailed investigation of kinematic parameters. Athletes have diverse training backgrounds and muscular imbalances that may result from asymmetric training loads or previous injuries. These factors are not merely theoretical but significantly influence biomechanical reactions during repeated movements^29,36. For example, modified scapular mechanics might act as a compensation mechanism for rotator cuff weaknesses or instabilities³⁷.

Examining an athlete’s training history extends beyond injury considerations. The demands of different sports lead to distinct movement patterns. For instance, the biomechanics of a baseball pitcher’s shoulder differ from a swimmer’s shoulder due to the specific demands of each sport³⁸. This variation underscores that a universal approach is insufficient for understanding shoulder biomechanics and reflects the complex adaptation of the shoulder complex.

Therefore, a thorough examination of shoulder motion kinematics is crucial. This includes an in-depth study of joint elements and a comprehensive assessment of factors influencing an athlete’s biomechanical profile. Understanding scapular tilt, extension, humeral elevation, and individual athlete characteristics is essential for interpreting the shoulder complex’s responses to repetitive movements and chronic shoulder pain³⁹.

Exploring subjective experiences in athletes is also vital. Investigating whether variations in lower trapezius muscle activation correlate with pain or discomfort during or after repetitive movements connects objective physiological data to athletes’ experiences. This approach aligns with the biopsychosocial model, emphasizing the interplay between physiological factors, psychological experiences, and social influences in chronic shoulder pain⁴⁰. Evaluating the impact of lower trapezius muscle activity on scapular kinematics and shoulder stability could guide targeted rehabilitation strategies for neuromuscular imbalances associated with chronic shoulder pain⁴¹.

Integrating biomechanical analyses, such as motion capture and 3D kinematics, with EMG data offers a comprehensive view of how changes in lower trapezius muscle activation affect shoulder dynamics. This coordinated approach bridges functional outputs and muscle activity, providing detailed insights for developing interventions that address both broader movement patterns and specific muscles linked to chronic shoulder pain⁴².

In summary, examining electromyography of the anterior deltoid and lower trapezius muscles goes beyond mere statistical significance. It involves understanding the temporal characteristics of muscle activation, considering athletes’ subjective experiences, and exploring broader biomechanical implications for shoulder function. This holistic perspective enhances our understanding of adaptive strategies and neuromuscular imbalances associated with chronic shoulder pain, informing targeted rehabilitation strategies for optimal outcomes.

Despite the lack of statistically significant differences in shoulder and elbow coordination, analyzing specific joint coupling patterns reveals deeper insights into the adaptive motor control strategies of athletes, particularly those with chronic shoulder pain. Advanced analytical techniques, such as phase analysis, can uncover subtle changes in the temporal relationships between shoulder and elbow movements. This approach allows for a more detailed examination of joint coordination⁴³.

Further studies may reveal altered movement patterns in athletes, potentially indicating compensatory mechanisms that enable effective coordination despite persistent chronic shoulder pain. Athletes may adapt their joint movements in range, timing, or sequence to respond dynamically to fatigue and pain⁴⁴.

Combining biomechanical analyses with coordination data enriches our understanding of how changes in coordination, muscle activation, and joint kinematics interact. This integrated approach helps connect alterations in movement patterns with overall upper limb dynamics, providing a comprehensive view of how changes affect upper limb function⁴⁵.

Exploring intersegmental coordination, particularly in shoulder-elbow and elbow-wrist interactions, requires a nuanced analysis that transcends statistical significance. Employing advanced analytical techniques and focusing on underlying mechanisms will yield deeper insights into the adaptive strategies of athletes with chronic shoulder pain. This knowledge not only contributes to theoretical frameworks of motor control but also informs targeted interventions aimed at maintaining and optimizing upper limb function in the face of repetitive movement fatigue and chronic shoulder pain.

It should be noted that the results of this study can be criticized from some perspectives. It should be noted that factors such as pain intensity, pain duration, or prior physical activity levels may alter movement strategies and electromyographic responses. However, due to the limited number of subjects, it was not possible to analyze the possible effect of these factors on this study’s results.

Conclusion

Our study’s findings reveal that both groups were able to complete the repetitive movement task, but they employed distinct movement strategies. The CSP (Chronic Shoulder Pain) group primarily focused on controlling the shoulder joint, whereas the CON (Control) group utilized both the shoulder and elbow joints. This indicates that individuals with chronic shoulder pain may adopt a strategy that minimizes pain by relying more on non-painful joints, even if this results in a reduced duration of task performance. The observed difference in movement strategies between the groups highlights how chronic pain can lead to the development of compensatory mechanisms that aim to reduce discomfort and maintain functionality, albeit with some trade-offs.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

RS, SA, MS, and MA contributed to the study design and data collection. SA, RS, and MS drafted the manuscript and made critical revisions to the manuscript. All authors read and approved the final manuscript.

Funding

This work is based upon research funded by Iran National Science Foundation (INSF) under project no. 4013596.

Data availability

All data generated or analyzed during this study are included in supplementary information file 1.

Declarations

Ethics approval and consent to participate

Prior to starting the investigation, study approval was obtained from the Biomedical Research Ethics Committee of Allameh Tabatab’i University (Ethics code: IR.ATU.REC.1401.084), and all participants gave written informed consent. The authors confirm that all methods were performed in accordance with the relevant guidelines and regulations. Moreover, informed consent has been obtained to publish the images in an online open-access publication.

Consent for publication

Written consent for publication has been obtained from the patient, as shown in Fig. 2.

Competing interests

The authors declare no competing interests.