The influence of hand position on scapular kinematics in push-ups: comparing athletes with chronic shoulder pain and healthy controls

Abstract

Background

Scapulothoracic motion during arm elevation involves scapular posterior tilt (PT), upward rotation (UR), and external rotation (ER). Abnormal scapular kinematics are common in people with chronic shoulder pain, potentially exacerbating symptoms and impairing function. Push-ups, a common exercise for shoulder rehabilitation, may influence scapular motion but have not been extensively studied in this context. Thus, this study aimed to investigate the effect of hand position on scapular kinematics during a push-up exercise in athletes with and without chronic shoulder pain (CSP).

Methods

Twenty-four male overhead athletes were allocated into two groups: CSP (n = 12) and Control (CON; n = 12). Scapular kinematics in three planes (PT, UR, and ER) were measured using a Vicon motion capture system during push-ups in three hand positions (internal rotation, IR; neutral rotation, NR; and external rotation, ER). Measurements were taken in the concentric phase of the push-up. Statistical analyses using repeated-measures ANOVA assessed the effects of hand position and elbow extension on scapular kinematics between the two groups.

Results

For PT of the scapula, in the IR hand position, participants from both the CON and CSP groups showed similar decreases (CON group = from 25.74° to -16.10°; P < 0.001). In the NR hand condition, the CON group decreased PT from 16.12° to -15.98° (P < 0.001), but there was no significant change in the CSP group. In the ER hand condition, for the CON group, PT decreased from 18.63° to -9.38° (P < 0.026), with no significant change observed in the CSP group. For UR of the scapula, the CON group showed significant decreases in the IR hand condition (from 15.37° to -2.28°; P < 0.019), while the CSP showed minimal changes. In the IR hand condition, the ER of the scapula increased from 20.44° to 25.13° (P < 0.003) in the CON group. At the same time, the CSP showed smaller changes. In the NR hand condition, ER of the scapula in the CON group decreased from 24.79° to 9.38° (P < 0.001), with no significant change observed in the CSP group.

Conclusion

Scapular kinematics (UR, PT, and ER) differed significantly across hand rotation conditions and groups. The CON group exhibited more pronounced changes in these kinematic measures, while the CSP group showed limited variation. This may indicate an association between chronic pain and movement restriction. These findings emphasize the need for targeted rehabilitation strategies that consider these kinematic differences.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-025-06212-1.

Introduction

For athletes participating in overhead sports, shoulder pain is one of the most prevalent musculoskeletal problems and can be debilitating [1–4]. Indeed, sports like volleyball, baseball, softball, handball, tennis, and swimming that require an overhead motion combine high external rotational forces with quick speed, placing a great deal of biomechanical strain on the shoulder and increasing the risk of injury [5]. Unlike other major ball-and-socket joints such as the hip, the shoulder’s anatomical design prioritizes motion flexibility above stability [6]. Meanwhile, the scapula plays several key roles in facilitating shoulder function [7]. When a healthy shoulder is elevated in the scapular plane, the scapula shows a distinctive pattern of increased posterior tilt (PT), upward rotation (UR), and external rotation (ER) [8]. Also, the coordinated activation of the trapezius and serratus anterior muscles is critical for both the mobility and stability of the scapula during upper limb movements. These muscles function synergistically to maintain scapular stability on the thorax and rotations in all three degrees of freedom [9, 10].

Furthermore, understanding the scapula’s resting position and how it moves during exercises like push-ups is essential for shoulder rehabilitation and function. At rest, the scapula has a 10° anterior inclination and 35° internal rotation, positioning it horizontally [11]. Deviations from these patterns, known as scapular dyskinesis, may be present in those with shoulder injuries [7]. Many shoulder injuries in overhead athletes may be associated with scapular dyskinesis and a loss of control over resting and dynamic scapular posture [12, 13].

Restoring appropriate scapular kinematics during scapulohumeral motion is one goal of rehabilitation treatments for shoulder complex injuries. To achieve this, practitioners choose exercises that target the scapular-stabilizing muscles and position the scapula to maximize the subacromial space [8]. The push-up and its multiple variants are popular because they are easily adjusted to different difficulty levels and are thought to enhance joint stability and proprioception during performance due to joint compression forces [14]. Muscle activation and scapular kinematics may change depending on the hand position adopted when performing push-ups in strength training or rehabilitation settings [15, 16]. In a standard push-up position at approximately 90° shoulder flexion, healthy individuals typically exhibit 10–15° PT, 20–30° UR, and 25–35° internal rotation. Deviations from these norms in athletes with shoulder pain may reflect underlying pathology or compensation [17, 18]. Previous findings indicate that the anterior tilt of the scapula persists during the movement, revealing limited engagement of the posterior tilting mechanisms. Upward rotation remains present, demonstrating a consistent elevation pattern of the scapula, and internal rotation is maintained throughout the push-up motion without shifting toward an externally rotated position [8]. Similarly, it has been reported that the scapula remains anteriorly tilted and internally rotated during the concentric phase, reinforcing the concept that push-ups primarily engage scapular protraction and stabilization rather than retraction and posterior tilting [17, 18]. Improvements in shoulder stability and proprioception during push-ups result from dynamic muscle activation and neuromuscular control, not joint compression, which primarily affects joint congruence. Despite their clinical relevance, the effects of hand position on scapular kinematics in athletes with shoulder pain, particularly concerning shoulder ROM, remain underexplored [19, 20].

Push-ups, however, can be associated with shoulder impingement if the shoulders are raised [8]. Additionally, muscle fatigue can change scapular kinematics, especially in the case of the serratus anterior muscle, resulting in increased internal rotation and decreased posterior tilt of the scapula [21]. Therefore, it must be emphasized that the importance of better understanding the relationship between shoulder kinematics and muscle activation when considering exercise prescription and progression in shoulder rehabilitation.

On the other hand, it should also be kept in mind that the position of the hands affects scapular kinematics during push-ups [22]; it is worth considering if certain hand positions may predispose individuals to clinically detectable scapular abnormalities. For example, an internally rotated hand position may enhance scapular internal rotation and anterior tilt, increasing winging in people with neuromuscular dysfunction [21]. An externally rotated hand position, on the other hand, may increase serratus anterior muscle activation, which is necessary for scapular stability and preventing excessive internal rotation of the scapula [15]. These potential consequences emphasize the need to assess push-up variations to optimize techniques for those with and without shoulder pain.

To our knowledge, the implications of different hand positions on scapular kinematics are still unclear. As push-ups require significant shoulder elevation and can aggravate symptoms in those with shoulder pain, studying scapular kinematics holds clinical relevance [23]. Furthermore, most research has focused on healthy people, leaving a gap in our understanding of how scapular kinematics differ between athletes with and without chronic shoulder pain during functional exercises like push-ups. Examining variations in scapular kinematics within each group across different push-up conditions can provide deeper insight into movement adaptations. Consequently, our study examined the impact of hand position on scapular kinematics during push-ups and compared scapular motion in athletes with and without shoulder pain.

Methodology

Participants

In this cross-sectional study, a convenience sample of 24 male overhead athletes from ball sports (including volleyball, basketball, and handball) was recruited and allocated into two groups: a group with chronic shoulder pain (CSP) (12 individuals) and the control group (CON) (12 individuals). Power analysis was used to estimate the sample size using GPower software version 3.1. According to a previous study [24] and considering changes in scapular cross-flexion movement due to fatigue on the second day while performing shoulder internal rotation movement, and considering an α = 0.05 and β = 0.20, the sample size was 21 participants. In this study, 24 participants were considered to reduce the possible effect of sample dropout. For the CSP group, the athletes had to meet the following criteria: a history of lateral shoulder pain for a minimum of 3 months on their dominant upper limb (the arm used to throw a ball), no pain radiation to the cervical or distal segments of the upper extremity, and no history of fever, night pain, or cold sweats during the pain period. The control group consisted of overhead athletes with at least 3 years of regular activity in a sport requiring overhead movements and no history of shoulder pain lasting more than a week.

Overhead athletes with a self-reported history of cervical radiculopathy, glenoid labrum lesions, shoulder dislocation, shoulder girdle muscle rupture, a history of neurological, vestibular, or other balance-affecting medical conditions, and those using any medications for neurological or metabolic disorders were excluded. An experienced musculoskeletal physical therapist conducted clinical assessments, including active and passive shoulder ROM (flexion, internal/external rotation) and thoracic mobility (extension/rotation), to confirm no significant limitations (< 10% deficit compared to normative values) to exclude athletes with possible shoulder ROM deficits [25]. Moreover, the mentioned physiotherapist performed neurological examinations (radiating pain/paresthesia pattern, head/neck extension test, significant muscle atrophies in the upper extremities) were performed to rule out cervical radiculopathies. The New York Posture Rating Scale was used to exclude participants with significant postural deformities such as hyperkyphosis, scoliosis, or forward head posture. The right hand was dominant in all of the participants. The Biomedical Research Ethics Committee of Allameh Tabataba’i University (ATU) (Ethics code: IR.ATU.REC.1401.084) approved this study.

Instrumentation

Three-dimensional whole-body posture and movement characteristics were recorded with a sampling frequency of 120 Hz using a high-resolution six-camera Vicon MX3 motion capture system (Vicon Peak, Oxford Metrics Ltd., Oxford, UK), utilizing passive reflective markers. These markers were affixed to anatomical landmarks with double-sided adhesive tape, according to the specifications of our linked-segment model. The body was segmented into 15 distinct portions: head (5 markers), trunk (5 markers), pelvis (5 markers), arms (3 markers per arm), forearms (4 markers per forearm), hands (3 markers per hand), thighs (3 markers per thigh), legs (3 markers per leg), and feet (3 markers per foot).

Scapular movements were analyzed as ER, UR, and PT relative to the thorax (Fig. 1).

Fig. 1.

Movements of the scapula, including (A) External/Internal rotation, (B) Upward/Downward rotation, and (C) Anterior/Posterior relative to the thorax

Testing procedures

All of the tests were completed in one session, and data were collected from the dominant upper limb.

Prior to the tests, the participants performed a ten-minute warm-up and stretching protocol. Following the warm-up, they practiced push-ups to achieve optimal performance. Push-ups were performed using three different hand positions (hands internally rotated 90 degrees, neutral, and hands externally rotated 90 degrees) (Fig. 2). All participants were allowed to perform fifteen trials (five for each position) to familiarize.

Fig. 2.

Palm position and corresponding output in (A) internal rotation, (B) neutral, and (C) external rotation conditions

Markings on the ground were used to help participants to place their hands at the correct angle relative to a reference line set perpendicular to the body. The internal, neutral, and external hand positions were defined by rotating the palms 90 degrees inward, maintaining a neutral position, and rotating 90 degrees outward from this reference line, respectively [26]. Participants with limited shoulder ROM were monitored to ensure no compensatory scapular motion occurred during hand positioning.

In each palm position condition, participants completed three repetitions of push-ups (two counts down, two counts up), following a manual count where the researcher verbally guided them through the timing of each repetition to ensure consistency across trials. Participants were instructed to make light contact with a designated block (placed beneath their chest) at the end of the eccentric phase and to fully extend their elbows at the end of the concentric phase for each repetition. The order of conditions was randomized using simple randomization via a random number table. To minimize fatigue effects, each condition was separated by at least one minute of rest. No participants reported fatigue following any of the trials.

Data reduction

To assess scapular kinematics during push-ups, a 3D motion capture system was employed, utilizing multiple high-speed cameras to accurately track the movements of reflective markers placed on specific anatomical landmarks. Markers were attached to the acromion, inferior angle, and vertebral margin of the scapula, and the levels of the spinous processes of the T1 and T8 vertebrae. This strategic placement allowed for a clear reference frame to measure scapular motion relative to the thorax throughout the push-up exercise [27]. The study focused on three primary variables: PT, UR, and ER. These variables were measured in 10° increments during the push-up motion, capturing the scapular mechanics while the upper limbs moved through the defined range of motion from maximum elbow flexion (105°) to minimum elbow extension (35°). This range was chosen based on its functional relevance in various athletic activities, ensuring that the analysis aligned with movements commonly performed in both sports and daily life [25].

Mean scapular rotations were averaged across three repetitions for each participant to ensure the reliability of the kinematic data [28]. The selection of three repetitions was based on standard practices in biomechanical research to balance data reliability with participant fatigue. This averaging process minimized the impact of individual variability and facilitated a more comprehensive analysis of scapular movement patterns. The data normalization techniques implemented included time normalization to maintain consistent speeds across trials, which is crucial for making valid comparisons between different conditions. Additionally, during analysis, angle normalization was utilized to standardize scapular angles, allowing researchers to observe trends in scapular motion across different push-up variations. These normalization techniques helped to ensure that the observed changes in scapular mechanics were attributable to variations in push-up conditions rather than speed or effort fluctuations [27].

Utilizing a low-pass Butterworth filter, the raw kinematic data were filtered, effectively reducing noise and enhancing the clarity of the observed movement patterns. This step is essential in motion analysis to ensure that the resulting data reflect true movement characteristics rather than artifacts introduced by the measurement system [29]. The subsequent analysis utilized Euler angles to quantify scapular motion in three-dimensional space, allowing for a detailed assessment of how various shoulder positions influenced scapular kinematics during the push-ups [8]. Euler angles provide a comprehensive means to interpret complex movements, enabling a better understanding of the three-dimensional orientation of the scapula [28]. This comprehensive methodological approach, which integrates precise anatomical tracking with robust analytical techniques, underscores the critical role of scapular mechanics in optimizing exercise performance and informing rehabilitation strategies for shoulder-related injuries.

Statistical analysis

SPSS (version 26.0) was used for statistical analysis. The Shapiro-Wilk test was used to examine the normal distribution of the data. One-way analyses of variance were used to examine for between-group differences in demographic variables. One-way repeated-measures analyses of variance (ANOVAs) were conducted to examine the effects of hand positions on scapular kinematics during push-ups between the two groups. Post hoc analyses with Bonferroni correction, implemented in SPSS to automatically adjust p-values for multiple comparisons (α = 0.05), were conducted to examine between-group differences.

Results

Characteristics of the participants

The demographic variables of the participants from the CSP and control groups are shown in Table 1. There were no significant between-group differences observed for the two groups.

Table 1.

Demographic characteristics of the participants

Variables	CSP (N = 12)	CON (N = 12)	P -value
	Mean ± SD	Mean ± SD
Age (years)	23.25 ± 2.76	21.66 ± 2.01	0.123
Body Mass (kg)	72.66 ± 8.74	78.50 ± 12.70	0.204
Height (cm)	182.58 ± 6.90	183.33 ± 12.82	0.860
BMI (kg.m2)	21.77 ± 2.08	23.33 ± 2.78	0.134
VAS (mm)	2.83 ± 0.88	N/A	-

Notes: †statistically significant differences between the CON and CSP groups. **P-value < 0.05 is considered to be statistically significant

Abbreviations: CON = Control Group; CSP = Athletes with Shoulder Chronic Pain Group; BMI = Body Mass Index; VAS = Visual Analogue Scale of current pain

Scapular kinematics

Table 2; Figs. 3, 4 and 5 show scapular UR, PT, and ER in 10° increments across the elbow ROM during the concentric phase of the push-up exercise for each hand position. To orient the reader, for the neutral (0-degree rotation) hand condition, the scapula exhibited a general pattern of decreasing PT and ER, while UR showed a variable response between groups.

Table 2.

Scapular posterior Tilt (PT), upward rotation (UR), and external rotation (ER) at 10° increments across the elbow extension range of motion ^a, for different hand positions

Variables			Elbow Flexion (°)								p-values (rotation)b	p-values (rotation × groups) b
			105 (°)	95 (°)	85 (°)	75 (°)	65 (°)	55 (°)	45 (°)	35 (°)
Cardinal Plane	Position		Mean ± SD
Palms Rotation	Posterior tilt of the Scapula
	Internal hand condition	CON	25.74 ± 2.94*	21.11 ± 2.40*	14.97 ± 1.95*	8.31 ± 1.59*	0.92 ± 1.46*	-6.90 ± 1.16*	-13.33 ± 0.89*	-16.10 ± 1.91*	P < 0.001**	P < 0.001**
		CSP	8.92 ± 1.19*	4.49 ± 0.88*	-2.08 ± 1.35*	-7.33 ± 1.47*	-11.66 ± 1.74*	-15.51 ± 2.14*	-18.10 ± 2.43*	-19.44 ± 2.30*
	Normal hand condition	CON	16.12 ± 2.40*	12.14 ± 2.04*	7.49 ± 1.95*	2.87 ± 2.18*	-3.60 ± 2.59*	-8.00 ± 3.10*	-12.18 ± 3.66*	-15.98 ± 4.17*	P < 0.001**	P < 0.082
		CSP	11.17 ± 2.69*	6.88 ± 2.65*	1.71 ± 2.42*	-2.17 ± 2.51*	-6.65 ± 2.81*	-10.49 ± 2.56*	-14.14 ± 2.48*	-12.08 ± 4.27*
	External hand condition	CON	18.63 ± 2.09*	14.93 ± 1.82	10.37 ± 1.65*	4.48 ± 2.20*	0.20 ± 2.57*	-4.23 ± 3.09*	-7.05 ± 3.30*	-9.38 ± 3.48*	P < 0.026**	P < 0.001**
		CSP	3.25 ± 0.84	1.12 ± 1.11	-0.66 ± 1.24*	-2.83 ± 1.33*	-4.26 ± 2.00	-8.01 ± 2.79*	-8.31 ± 3.22	-6.50 ± 4.29
	Upward rotation of the Scapula
	Internal hand condition	CON	15.37 ± 4.62	13.82 ± 4.57*	7.57 ± 5.17	5.24 ± 5.05	2.24 ± 5.09	-0.05 ± 4.97	-1.17 ± 5.03	-2.28 ± 5.06	P < 0.019**	P < 0.004**
		CSP	-6.32 ± 2.26	-6.48 ± 1.92	-7.64 ± 0.90	-7.82 ± 0.68	-7.51 ± 0.63	-7.20 ± 0.61	-6.56 ± 0.71	-5.91 ± 0.82
	Normal hand condition	CON	-27.54 ± 4.99	-24.97 ± 4.55	-21.79 ± 4.04*	-18.89 ± 3.55	-16.21 ± 3.15*	-14.13 ± 3.03*	-13.26 ± 2.91	-12.57 ± 2.94	P < 0.001**	P < 0.827
		CSP	2.40 ± 2.64*	4.79 ± 3.07	8.26 ± 3.19*	10.93 ± 3.06*	13.17 ± 2.73*	15.22 ± 2.62*	17.19 ± 2.59*	18.12 ± 2.47*
	External hand condition	CON	-3.53 ± 2.51*	-0.81 ± 2.49*	5.10 ± 2.55*	9.60 ± 2.40*	13.98 ± 2.52*	18.15 ± 2.76*	20.91 ± 2.87*	23.09 ± 3.00*	P < 0.001**	P < 0.425
		CSP	-2.69 ± 1.78*	0.59 ± 2.28	3.95 ± 2.51*	7.44 ± 2.90*	11.28 ± 3.09*	15.02 ± 3.31*	18.63 ± 3.24*	21.06 ± 3.21*
	External rotation of the Scapula
	Internal hand condition	CON	20.44 ± 1.45	21.60 ± 1.65	22.42 ± 2.04	22.99 ± 2.39*	23.08 ± 2.81*	23.94 ± 2.93*	24.58 ± 3.17*	25.13 ± 3.24*	P < 0.623	P < 0.003**
		CSP	15.58 ± 1.12	14.70 ± 1.30	13.64 ± 1.44	12.15 ± 1.59	10.98 ± 1.61*	10.21 ± 1.79*	8.91 ± 2.13*	8.65 ± 2.35
	Normal hand condition	CON	24.79 ± 2.91	20.69 ± 2.68	18.64 ± 2.68	16.76 ± 2.67	14.90 ± 2.66	13.20 ± 2.54	10.55 ± 2.39	9.38 ± 2.31	P < 0.001**	P < 0.019**
		CSP	7.50 ± 3.60	5.51 ± 3.07	3.87 ± 2.62	2.49 ± 2.21	1.69 ± 1.95	1.43 ± 1.79	1.70 ± 1.70	1.50 ± 1.72
	External hand condition	CON	9.01 ± 3.90	1.33 ± 3.32	-3.46 ± 3.02*	-6.77 ± 2.93*	-7.29 ± 3.33*	-10.30 ± 3.06*	-11.27 ± 3.25*	-12.07 ± 3.54*	P < 0.001**	P < 0.305
		CSP	6.57 ± 3.26*	4.77 ± 3.15*	1.37 ± 2.76*	-1.74 ± 2.31*	-4.22 ± 2.11*	-6.32 ± 2.03*	-7.90 ± 1.95*	-8.58 ± 1.96*

Note: ^a negative number indicates movement in the opposite direction; ^b P -values (rotation) represent the significance of changes in scapular movement across elbow flexion angles within each condition. P-values (rotation × groups) indicate the interaction effect, testing for between-group differences in these trends; ** shows a significant interaction effect for the ANOVA (p < 0.05); * Shows a significant effect within groups for the ANOVA (p < 0.05)

Abbreviations: CON = Control Group; CSP = Athletes with Shoulder Chronic Pain Group

Fig. 3.

Effect of different hand rotational positions (90° IR, 0° rotation, and 90° ER) on scapular PT across the range of elbow flexion (mean ± SD)

Fig. 4.

Effect of different hand rotational positions (90° IR, 0° rotation, and 90° ER) on scapular UR across the range of elbow flexion (mean ± SD)

Fig. 5.

Effect of different hand rotational positions (90° IR, 0° rotation, and 90° ER) on scapular ER across the range of elbow flexion (mean ± SD)

Posterior Tilt of the scapula

There were significant between group differences in scapular PT across different elbow flexion angles and hand rotation conditions (Table 2; Fig. 3). In both groups, PT decreased over time, shifting from a position of posterior tilt at the start of the concentric phase to anterior tilt by the end, as indicated by the line crossing the zero-degree PT line. This suggests that at the bottom of the push-up, the CON group positioned the scapula in greater posterior tilt, potentially allowing for a greater range of motion on the way up.

In the IR hand condition, the CON group exhibited a significant reduction in PT, decreasing from 25.74° at 105° elbow flexion to -16.10° at 35° (P < 0.001), while the CSP group showed a smaller range of change, from 8.92° to -19.44°. In the normal hand condition, the CON group’s PT decreased significantly over time, from 16.12° to -15.98° (P < 0.001), with a significant difference compared to the CSP group (F = 4.35, P < 0.033), which showed a less pronounced reduction from 11.17° to -12.08°. Under the ER hand condition, the CON group again showed a significant PT decrease, from 18.63° to -9.38° (P < 0.026), whereas the CSP group exhibited minimal changes, from 3.25° to -6.50°, with no significant differences over time (F = 4.69, P = 0.657).

Upward rotation of the scapula

Distinct between-group patterns of UR of the scapula were observed across different elbow flexion angles and hand rotation conditions (Table 2; Fig. 4). In the IR hand condition, the CON group exhibited a significant reduction in UR, from 15.37° at 105° elbow flexion to -2.28° at 35° (P < 0.019), while the CSP group remained relatively stable, with values ranging from − 6.32° to -5.91° (P < 0.004). In the neutral hand condition, the CON group showed a steady decrease in UR from − 27.54° at 105° elbow flexion to -12.57° at 35° (P < 0.001). In contrast, the CSP group gradually increased from 2.40° to 18.12°, but with no significant changes over time (P = 0.827). Under the ER hand condition, the CON group demonstrated a significant increase in UR, rising from − 3.53° at 105° elbow flexion to 23.09° at 35° (P < 0.001), while the CSP group exhibited a gradual increase from − 2.69° to 21.06° across the flexion angles, but with no significant changes (P = 0.425).

External rotation of the scapula

Distinct between-group patterns of ER of the scapula were also observed across different elbow flexion angles and hand rotation conditions (Table 2; Fig. 5). In the IR hand condition, the CON group showed a significant increase in ER of the Scapula, rising from 20.44° at 105° elbow flexion to 25.13° at 35° (P < 0.003). In contrast, the CSP group exhibited a non-significant decrease in ER of the Scapula, with values decreasing from 15.58° to 8.65° across the flexion angles. In the neutral hand condition, the CON group showed a significant decrease in ER of the Scapula from 24.79° at 105° elbow flexion to 9.38° at 35° (P < 0.001), whereas the CSP group also showed a gradual decline, from 7.50° to 1.50°, but with no significant changes (P = 0.019). In the ER hand condition, the CON group experienced a significant decrease in ER of the Scapula, from 9.01° at 105° elbow flexion to -12.07° at 35° (P < 0.001), while the CSP group showed a gradual decrease from 6.57° to -8.58°, but the changes were not significant (P = 0.305).

Discussion

This study examined the effect of hand position on scapular kinematics during the concentric phase of push-up exercises in male athletes with and without CSP. The study findings demonstrated significant differences in scapular PT, UR, and ER across three hand positions—IR, NR, and ER. The CON group showed greater changes in PT, UR, and ER across all hand positions. Particularly, more reduction in PT and UR was present in the CON group. In addition, the CSP group demonstrated limited kinematic variability while performing push-ups, which may indicate movement pattern alterations associated with chronic pain in the shoulder. These results highlight that altered kinematics in individuals with CSP may necessitate tailored rehabilitation strategies to address specific movement impairments.

Many push-up exercises in clinical settings are prescribed to correct scapular control. This idea is underpinned by reported higher activation of the serratus anterior muscle in comparison with the upper part of the trapezius muscle during push-ups [30]. Nevertheless, there is limited information regarding the scapular kinematic patterns during these exercises [11]. Changes in the kinematics of the scapula during arm motion have been previously associated with reported pain in the glenohumeral joint [31]. In line with this observation, existing literature has shown changes in the kinematics of the scapula at specific angles of arm elevation in people with symptomatic scapular dyskinesis (SD) [32, 33]. Furthermore, participants with SD showed increased IR and anterior tilt of the scapula when lowering the arm at 120, 90, 60, and 30 degrees compared with individuals without scapular dyskinesis [17, 29, 34]. However, to our knowledge, no previous studies have examined this issue in athletes with CSP.

Although evaluating specific scapular angles allows for identifying scapular movement changes associated with symptomatic shoulders, this method overlooks other movement features, such as temporal data and the time series profile throughout the entire scapular movement range [17, 35]. Important information may be missed via the discrete data approach by neglecting correlation and variability in biomechanical time series [36].

Although past studies have indicated a relationship between chronic shoulder pain and scapular movement patterns, it should be kept in mind that the relationship between shoulder symptoms and SD is controversial [37–39]. While a previous investigation indicated that the frequency of SD in overhead athletes is higher than in other athletes [40], there is no evidence to suggest that the prevalence of SD is higher in individuals with symptomatic shoulders when compared with asymptomatic individuals [41]. Furthermore, it has been reported that SD is not a risk factor for shoulder injury or pain development in baseball pitchers [42]. In contrast, overhead athletes with SD have shown a 43% rise in the possibility of shoulder pain [43]. Thus, it is unclear whether SD results from natural movement variability or is related to shoulder injury and pain. It would seem that it would be important to define normal and abnormal scapular movement, given the great degree of variation and the probability of flexible and optimized strategies [38, 39].

In the only other study on the kinematics of the scapula during push-up exercises [8], the authors reported similarities in the kinematics of the scapula during wall push-ups as observed in the current investigation. At the start of the concentric phase of wall push-ups and across all push-up positions examined, the scapula was in a forward rotated, UR, and IR position. In both studies, a reduction in ER was observed throughout the movement range. However, in the IR hand position in the current investigation, a different UR pattern was observed compared with the other two positions. In the IR hand condition, UR at the beginning of the elbow movement range was higher than in the neutral and ER hand conditions. Still, in the middle of the movement range, UR was higher in the IR than in the other two positions. These differences may relate to the activation of different muscles and the need for scapular stabilization in various degrees of elbow flexion.

Ultimately, the findings of the present study indicate that the concentric push-up exercise leads to significant changes in the PT, UR, and ER of the scapula. These changes may be related to the activation of different muscles and the need for scapular stabilization in various arm height positions. On the other hand, the observed changes may be due to differences in kinematic patterns between healthy participants and individuals with shoulder pain. Consequently, a more precise and detailed assessment of scapular kinematics during push-ups is recommended to better inform exercise selection for athletes with and without shoulder pain.

Although significant differences in scapular kinematics were observed between the CON and CSP groups, factors such as subtle ROM limitations, gender differences, scapular positioning, or undetected muscle imbalances may have influenced the results. The study controlled for major ROM deficits and postural deformities, but minor variations could still affect outcomes. Additionally, the absence of sEMG data limited insights into muscle activation patterns. Future studies should assess these factors, including detailed ROM evaluations, thoracic mobility, and muscle activity, to better understand their impact on scapular kinematics and inform rehabilitation strategies.

Conclusion

This study demonstrated that changes in hand position during push-up exercises can affect scapular positioning. Scapular kinematics (UR, PT, and ER) differed significantly across hand positions and groups. The CON group exhibited more pronounced changes in these kinematic measures, while the CSP group showed limited variation. The results of altered scapular kinematics in the CSP group may highlight the importance of conducting further studies to design more effective rehabilitation protocols.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CON

Control) group

CSP

Chronic shoulder pain

ER

External rotation

NR

Neutral rotation

PT

Posterior tilt

ROM

Range of motion

SD

Scapular dyskinesis)

UR

Upward rotation

Author contributions

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

Funding

None declared. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated and analyzed during the current study are available in Supplementary 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.

Competing interests

The authors declare no competing interests.