Scapular dyskinesis-based exercise therapy versus multimodal physical therapy for subacromial impingement syndrome in young overhead athletes with scapular dyskinesis: a randomized controlled trial

Abstract

Objective

To compare the efficacy and underlying mechanisms of scapular dyskinesis-based exercise therapy (SDBET) and multimodal physical therapy (MPT) in young male overhead athletes with subacromial impingement syndrome (SIS) and scapular dyskinesis (SD).

Methods

This single-center, single-blind, parallel-group superiority randomized controlled trial was designed to compare two interventions in athletes clinically diagnosed with SIS and SD. Outcome assessors were blinded to group allocation. Sample size was determined a priori using G*Power (f = 0.25, α = 0.05, power = 0.80), yielding a required sample of 28; 32 participants were planned to account for potential dropout. Inclusion criteria were: male overhead athletes aged 18–25 years with clinically diagnosed SIS and SD. Exclusion criteria included prior surgery or other contraindications. Participants were randomized to either SDBET or MPT groups. Both groups received 8 weeks of intervention (3 sessions/week) and a 4-week follow-up. The primary variable was disability, measured using the Shoulder Pain and Disability Index (SPADI). The secondary variables were pain (measured using the Visual Analog Scale [VAS]), shoulder active range of motion (AROM, measured using goniometry), strength (measured using Isometric Strength Tests [IST] of scapular stabilizers and external rotators), and scapular kinematics (measured using the Scapular Dyskinesis Test [SDT]). Assessments were conducted at baseline, week 8, and week 12.

Result

A total of 32 participants were analyzed (SDBET group: n = 16; MPT group: n = 16). Mixed-design ANOVA revealed significant main effects of time for all outcomes (all p < 0.001, η² = 0.71–0.92), and significant group effects for SPADI and IST (p < 0.01). Significant Time × Group interactions were also observed for all variables (p < 0.05, η² = 0.15–0.76), prompting further simple effects analysis. Disability (SPADI) decreased significantly in both groups by week 8 (p < 0.001, η² = 0.03), with no between-group difference. However, only the SDBET group maintained improvements at week 12 (p < 0.001, η² = 0.59). Pain (VAS) decreased more in the MPT group at week 8 (p = 0.018, η² = 0.17), but rebounded by week 12, eliminating group differences (p = 0.268, η² = 0.04). Active range of motion (AROM) improved in both groups by week 8 (p < 0.001), with only the SDBET group sustaining these gains at week 12 (p < 0.001, η² = 0.37). Strength (IST of scapular stabilizers and external rotators) improved exclusively in the SDBET group at both week 8 and 12 (all p < 0.001), with significant between-group differences (η² = 0.57–0.74). Scapular kinematics (SDT) improved in 43.8% of SDBET participants, while no change was observed in the MPT group (p = 0.001–0.004).

Conclusion

Both SDBET and MPT improved shoulder disability, pain and AROM in SIS patients with SD, but only SDBET showed lasting effects. Targeted therapy has more comprehensive and sustained effects.

Trial registration

This study was retrospectively registered in the Chinese Clinical Trial Registry on 03/04/2025 (Registration Number: ChiCTR2500100213), after the completion of participant enrolment due to administrative oversight. Although registration occurred post-enrolment, the study protocol was finalized and ethically approved prior to participant recruitment. The authors acknowledge the importance of prospective trial registration and are committed to adhering to this standard in all future trials.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13102-025-01254-8.

Introduction

Subacromial impingement syndrome (SIS) is the most common clinical diagnosis among patients with shoulder pain, accounting for approximately 44–65% of all shoulder pain cases [1]. The concept of SIS was first introduced by Neer in 1972 [2] and is characterized by compression of soft tissues, such as the supraspinatus tendon, subacromial bursa, and the long head of the biceps tendon, beneath the acromion during shoulder movement. Clinically, SIS is typically diagnosed using the Neer test, Hawkins test, and Jobe test. The underlying causes of SIS can be classified as either primary or secondary impingement [3, 4]. Primary impingement results from structural abnormalities or degenerative changes within the subacromial space [3, 5], such as abnormal acromial morphology and subacromial osteophytes [2]. Secondary impingement, on the other hand, arises from functional abnormalities in shoulder movement [4], including muscle imbalances [6] and muscle fatigue [7].

The prevalence of SIS is notably high among overhead athletes [8, 9], accounting for 27% of all shoulder injuries in this population [10]. Reports indicate that the prevalence of SIS among collegiate swimmers reaches 47% [11], significantly impairing their athletic performance and competition outcomes. Notably, the shoulder injuries experienced by these athletes are predominantly caused by secondary impingement [12, 13], which worsens during arm elevation and overhead throwing activities [14]. In overhead sports, scapular dyskinesis has been identified as a key factor contributing to the development of impingement symptoms [8, 15–17]. During arm elevation, inadequate scapular upward rotation, posterior tilt, and external rotation may result in impingement [18, 19]. Persistent abnormal scapular movement increases friction between the joint capsule, rotator cuff, and coracoacromial arch, leading to inflammation, degeneration, and even tears of the rotator cuff [20].

Scapular dyskinesis (SD) is defined as an abnormal position or movement of the scapula at rest or during motion, which may include excessive mobility, asymmetry, or scapular winging [15]. SD is not classified as a distinct disease but is frequently associated with other shoulder disorders. These alterations may contribute to or accompany changes in the subacromial space, thereby increasing the risk of shoulder injuries [21]. SIS and SD exhibit similar scapular kinematic alterations [16], such as reduced scapular upward rotation, increased internal rotation, and excessive protraction [1], which reduce the subacromial space and compress the shoulder’s soft tissue structures. These kinematic changes are not only linked to abnormal activation of the scapular muscles [22–24], but also to imbalances in muscle length and strength [23, 25, 26]. Muscle activation abnormalities in SIS patients include increased upper trapezius (UT) activity, inhibited activation of the middle trapezius (MT), lower trapezius (LT), and serratus anterior (SA), as well as an imbalance in the UT/LT activation ratio [22]. Muscle length and strength imbalances manifest as tightness and shortening of the levator scapulae (LS), UT, teres major (TM), and pectoralis minor (PM), while the MT, LT, SA, and rhomboids (RM) are weakened and elongated [27, 28]. This emphasizes the necessity of a comprehensive physical assessment to identify the presence of SD in SIS patients and implement targeted treatment strategies accordingly.

Approximately 90–95% of SIS cases are managed conservatively [29], with physical therapy and exercise therapy serving as the primary treatment modalities. Traditionally, physical therapy aimed at relieving pain and improving joint mobility—such as ultrasound therapy, manual therapy, and range of motion exercises—has been commonly employed to address shoulder disorders [30, 31]. In recent years, exercise interventions combining open kinetic chain (OKC) and closed kinetic chain (CKC) strategies have gained increasing attention in the early rehabilitation of SIS [32]. While CKC exercises have been extensively utilized in lower limb rehabilitation [33], growing evidence supports their effectiveness in shoulder and rotator cuff rehabilitation. These exercises have been shown to enhance shoulder joint stability and proprioception [34, 35].

However, evidence on effective treatments for SIS with SD is limited. Few studies have assessed SD before treatment and compared different therapies in this population. Although some SD-focused interventions have been reported [36, 37], overall evidence remains insufficient. This study aimed to assess SD in overhead athletes with SIS and compare the effectiveness of scapular dyskinesis-based exercise therapy (SDBET) and multimodal physical therapy (MPT). It also explored possible mechanisms behind differences in treatment outcomes. The primary hypothesis of this study was that SDBET would be superior to MPT in improving clinical outcomes at 12 weeks in patients with SIS accompanied by SD.

Methods

Study design

This study was designed as a single-center, single-blind, parallel-group randomized controlled trial (RCT) with a 1:1 allocation ratio, aiming to evaluate the effects of Scapular Dyskinesia-Based Exercise Therapy (SDBET) compared to Multimodal Physical Therapy (MPT) on subacromial impingement syndrome (SIS) in young overhead athletes with scapular dyskinesis (SD). Participants were randomly assigned using a computer-generated sequence, outcome assessors and data analysts were blinded to group allocation. The intervention lasted 8 weeks, with outcomes measured at baseline, post-intervention, and follow-up. Participants were recruited from a sports academy, a sports club, and a rehabilitation department of a hospital in Tongliao, China. The intervention for the SDBET group was delivered in a sports academy, while the MPT group received physical therapy at a local rehabilitation hospital. Adherence to the intervention was tracked using session attendance records. The study timeline was as follows:

Recruitment period: september 28, 2024– october 15, 2024

During this phase, participants were identified, screened, and randomly assigned to either the SDBET group or the MPT group. All participants completed the informed consent process and underwent baseline assessments, including data collection for all outcome measures.

Intervention period: october 20, 2024– december 16, 2024

Participants in the SDBET and MPT groups received their respective interventions over an eight-week period. At the end of the intervention, all participants completed post-intervention assessments for the designated outcome measures.

Follow-up period: december 17, 2024– january 14, 2025

After completing the intervention, all participants resumed their regular overhead sports training. A four-week follow-up assessment was conducted to evaluate the longer-term effects of the interventions, and final data collection was completed at the end of this period.

This study adhered strictly to the CONSORT guidelines for reporting randomization and allocation processes, and a detailed flowchart illustrating the study procedures is provided (see Fig. 1).

Fig. 1.

CONSORT Flow Chart of Participant Recruitment, Allocation, Intervention, and Follow-Up

Participants

Sample size calculation

This study utilized a repeated-measures design, with sample size estimated using G*Power 3.1.9.2. The significance level (α) was set at 0.05, statistical power (1 − β) at 0.80, and the effect size (f) at 0.25, corresponding to a moderate effect size as defined by Cohen [38]. This estimate was informed by previous studies investigating exercise-based interventions for shoulder pathologies [39] and is consistent with the established minimal clinically important difference (MCID) for the primary outcome SPADI, which ranges from 14.1 to 20.6 points [40].

Two groups (SDBET and MPT) were assessed at three time points (baseline, Week 8, and Week 12). The correlation among repeated measures was set at 0.5, and a non-sphericity correction factor of 1 was applied, assuming sphericity. Using the “ANOVA: repeated measures, within-between interaction” model, the required total sample size was 28 (14 per group). To account for potential attrition, a 15% increase was applied, resulting in a final sample size of 32 (16 per group), which was sufficient to maintain the desired statistical power. To address the risk of type I error due to multiple comparisons across time points and outcomes, Bonferroni correction was applied in all post hoc analyses.

Inclusion and exclusion criteria

Participants were screened for SIS based on Cools et al. [3] and for SD following Kibler et al. [41]. Inclusion criteria:

1. Male overhead athletes training ≥ 3 times per week.

2. Positive in at least two of Hawkins-Kennedy, Neer, and Jobe (empty can) tests, with anterior shoulder pain in the Apprehension test, to confirm SIS.

3. Negative full can test to exclude rotator cuff tears.

4. Negative Speed and O’Brien tests to rule out biceps tendinopathy.

5. Positive scapular dyskinesis Test (SDT).

6. Voluntary participation and adherence to study protocol.

Exclusion criteria: Acute shoulder injuries or inflammation, history of shoulder dislocation or fractures, severe range of motion limitations (e.g., inability to abduct or internally rotate), Recent shoulder surgery (e.g., rotator cuff repair, arthroscopy), cardiovascular disease or other contraindications. Additionally, Based on clinical experience, to avoid potential risks associated with the intervention, participants with a VAS score > 6 during active movement were excluded.

Recruitment and screening

Participants were recruited using purposive sampling based on predefined inclusion and exclusion criteria targeting male overhead athletes clinically diagnosed with SIS and SD. Recruitment was conducted via multiple channels, including poster advertisements, school-wide announcements, and referrals from healthcare professionals. Diagnoses were confirmed by two certified sports rehabilitation specialists and one licensed physical therapist, each with over 10 years of clinical experience. In China, sports rehabilitation specialists are trained in athletic injury assessment and corrective exercise, whereas physical therapists focus more on general physical dysfunction and clinical rehabilitation. To control for sex-related influences, only male participants were included.

All Participants provided informed consent after receiving full study details. They were informed of their right to withdraw from the study at any time without any repercussions. The study was approved by the Ethics Committee of Tongliao Mongolian Medicine Hospital (Approval No.: [TLSMYYY-2024-3-002]) and adhered to the Declaration of Helsinki (2013). Potential harms in this study included muscle soreness in the SDBET group and traction-related discomfort or joint tension in the MPT group. Improper execution in either group could provoke subacromial impingement pain. All interventions were closely supervised. When discomfort occurred, the affected exercise was immediately paused and modified (e.g., reduced range of motion or resistance), or the training volume was temporarily decreased. Symptoms were monitored, and participants resumed normal training once symptoms resolved.

Before randomization, baseline assessments included demographic data (age, height, weight) and SIS- and SD-related measures: shoulder pain and disability index (SPADI), visual analog scale (VAS) for pain in the Hawkins-Kennedy test, scapular dyskinesis test (SDT), active range of motion (AROM), isometric strength test (IST).

Randomization

To minimize selection bias, participants were randomly assigned using a computer-generated random number sequence. The randomization sequence was generated in SPSS 27 (IBM, Armonk, NY, USA) using the uniform distribution function, with values sorted in ascending order. The 16 participants with the smallest random values were assigned to the SDBET group, while the remaining 16 were allocated to the MPT group. Simple randomization without block or stratification was applied.

Opaque envelopes were sequentially numbered and sealed by a research coordinator not involved in recruitment or assessment. Each envelope was signed and dated to ensure concealment. Allocation concealment was verified by ensuring that envelopes were opened only after the completion of baseline assessments and in the presence of the principal investigator. A randomization log was maintained to document the integrity of the sequence. Randomization and group assignment were performed by an independent statistician who was not involved in participant recruitment, assessment, or intervention. Participants were informed of their group assignment only after baseline assessments were completed. Outcome assessors were blinded to group allocation throughout all measurement time points, resulting in a single-blind design. Baseline equivalence was assessed using independent samples t-tests for continuous variables and Fisher’s exact test for categorical variables.

Intervention

SDBET group intervention program

The SDBET exercise program was divided into three phases, with different interventions implemented at each stage. An individualized exercise dosage was applied. The SDBET group used the Borg CR10 scale to monitor exercise intensity, which was maintained at light to moderate levels to reduce the risk of injury. Specifically, intensity was kept between 2 and 3 during Phase 1, 3–4 during Phase 2, and 4–5 during Phase 3. The specific intervention protocol for the SDBET group is detailed in Table 1.

Table 1.

SDBET group intervention program

Phase	Objective	Content and Schedule	Specific Exercises and Execution
Phase 1	Activation of the scapular muscles and improvement of scapular movement control	Scapular Control Exercise (Week 1, total duration: 30 min)	1. Scapular Orientation Exercise) 2. Y-T-W-L Exercises 3.  Quadruped Scapular Protraction and Retraction
Phase 2	Release of tight muscles, strengthening of scapular and rotator cuff muscles, and enhancement of scapular movement control	Scapular Control Exercise + Scapular Stabilizer Exercise + Rotator Cuff strengthening (Weeks 2–6, session duration: 60 min)	In addition to Phase 1: 1.Scapular Stabilizer Exercise (stretching + strengthening) 2. Rotator Cuff strengthening exercises
Phase 3	Further reinforcement and maintenance of scapular motor control and stability	Exercise progression was implemented only when participants completed the previous exercises without reporting any serious pain or discomfort	In addition to Phase 2: 1.Lawnmower 2. Scapular Push-Ups 3. Plank 4. Internal and External Rotation at 90° Shoulder Abduction

Phase 1: scapular control exercises

The scapular control exercises consists of three exercises, each performed for three sets of 12–15 repetitions. The exercises are illustrated in Fig. 2.

Fig. 2.

Scapular Control Exercises

Scapular orientation exercise

Position: A therapist provides manual contact and verbal cues to guide participants in positioning the scapula optimally [42]. The following guidelines are followed: Visually, the scapula should be located between the second and seventh ribs on the posterior thorax [43], while tactile assessment involves palpating the spine and scapula to evaluate their relative positions [44]. The scapula is not aligned directly in the frontal plane but is oriented 30° anteriorly to the thoracic cage [45].

Movement: To restore normal scapular motion patterns, the therapist facilitates active scapular movements through manual guidance and verbal instructions, including protraction and retraction, abduction and adduction, elevation and depression, upward and downward rotation, and internal and external rotation [46, 47].

Y-T-W-L exercises

Relevant studies have demonstrated that these exercises effectively enhance the activation of both the scapular and rotator cuff muscles [48]. The exercises are divided into the following four variations:

Y exercise

Position: Prone on a fitness ball or treatment table, with elbows straight and 90° of shoulder flexion.

Movement: Slowly move the arms into a 45° angle relative to the body and then flex shoulder until they form a “Y” shape with the body, with thumbs pointing upward. Focus on contracting the lower trapezius, avoiding shoulder elevation or compensation.

T exercise

Position: Prone on a fitness ball or treatment table, with elbows straight and 90° of shoulder flexion.

Movement: Retract the scapula, positioning the arms to form a “T” shape with the body. Focus on contracting the middle trapezius, avoiding shoulder elevation or compensation.

W exercise

Position: Prone on a fitness ball or treatment table, with elbows bent at 90° and arms positioned at a certain angle forward, with forearms parallel to the chest.

Movement: Retract the scapula posteriorly and inferiorly, forming a “W” shape with the arms relative to the body. Focus on contracting the middle and lower trapezius.

L exercise

Position: Prone on a fitness ball or treatment table, with elbows bent at 90° and shoulder abducted to 90°.

Movement: Externally rotate the forearms (shoulder external rotation) and focus on contracting the rotator cuff muscles.

Quadruped scapular Protraction and Retraction

Position: Begin in a quadruped position with hands under the shoulders and knees under the hips. Toes gently touch the floor for balance. Maintain a neutral spine and engage the core to avoid lumbar extension or thoracic flexion.

Movement: Protraction: Push the shoulders forward, allowing the scapulae to glide laterally along the rib cage. Keep arms extended and feel the serratus anterior activation. Retraction: Draw the shoulders backward, bringing the scapulae toward the spine. Focus on engaging the rhomboids and middle trapezius. This is a closed kinetic chain exercise, as the hands remain in fixed contact with the ground, promoting scapular stability and proprioception.

Phase 2: scapular movement control exercises + scapular stabilizing exercises + rotator cuff exercises

Scapular stabilizing exercises (stretching + strengthening)

Based on the intervention strategies proposed by Lan Tang et al. [49], modifications were made to the exercises to improve their effectiveness and target specificity. Participants performed scapular stabilizing exercise tailored to the type of Scapular Dyskinesis (SD) they presented, ensuring more efficient and focused exercise. The strategies for scapular stabilizing exercise are outlined in Table 2.

Table 2.

Scapular stabilizing exercises strategies

Type	Characteristics	Intervention Strategy
Type 1	Scapular inferior angle protrusion, accompanied by scapular tilting in the sagittal plane.	Stretch the pectoralis minor; strengthen the serratus anterior, middle trapezius, and lower trapezius.
Type 2	Entire scapular medial border protrusion, accompanied by scapular internal rotation.	Stretch the teres major; strengthen the serratus anterior, middle trapezius, and rhomboid muscles.
Type 3	Scapular superior angle protrusion, accompanied by scapular elevation.	Stretch the upper trapezius and levator scapulae; strengthen the middle trapezius and lower trapezius.

Patients with different types of SD should select specific exercises based on the muscle that require stretching and strengthening, as outlined in the Table 3. Strengthening Exercise are illustrated in Fig. 3.

Table 3.

Implement of scapular stabilizing exercises (Stretch and strengthen)

Category	Muscle	Exercise	Implementation	Load
	Pectoralis Minor	Door Frame Stretch	Place the right hand or forearm against the door frame with the elbow slightly above shoulder height. Step one foot forward and push the body forward.	Perform each stretch 3 times, holding for 20 s each.
Stretch	Teres major	Wall Stretch	Place the palm on the wall above shoulder height, ensuring the arm is slightly bent. Gently lean the body forward, particularly feeling the stretch in the teres major area at the back of the shoulder.
	Upper Trapezius and Levator Scapulae	Neck Stretch	Use the left hand to hold the head while the right hand stabilizes the bench. Gently pull the head to the left and slightly rotate it to the right.
	Serratus Anterior	Foam Roller Wall Slide	Place both forearms shoulder-width apart on a foam roller at eye level against the wall. Slide the roller upward while keeping forearms in contact with the wall, focusing on serratus anterior contraction and scapular upward rotation. Then lower with control, avoiding shoulder shrugging.	Perform 10–12 repetitions per set, completing 3 sets.
		Resistance Band Scapular Protraction	The midsection of the resistance band is fixed to the back, with both hands holding the ends, shoulders abducted, and elbows slightly flexed. Push both hands forward, protracting the scapula, hold for 1 s, and feel the contraction of the serratus anterior.
		Wall Push-Up	Stand facing the wall, lean the body forward, and place both hands slightly wider than shoulder-width apart on the wall with elbows extended but not locked. Retract the scapula to bring the upper body closer to the wall. Engage the serratus anterior to protract the scapula, pushing the upper body away from the wall.
Strengthen	Middle Trapezius and Rhomboid	Wide-Grip Resistance Band Row	Fix the middle of the resistance band, hold the ends with both hands wider than shoulder-width apart, retract the scapula, and bend the elbows. Pull the resistance band towards the sides of the body, slightly wider than shoulder-width.
		Prone Horizontal Abduction and External Rotation	Prone on a fitness ball or treatment table, holding a dumbbell with the shoulder in 90° flexion in the sagittal plane. Then, perform shoulder horizontal abduction and external rotation, keeping the elbow fully extended throughout the movement [50].
	Lower Trapezius	Side-Lying External Rotation	Lie on your side on a treatment table, holding a dumbbell with the shoulder in a neutral position and the elbow flexed at 90°, with a towel placed between the elbow and the torso. Perform shoulder external rotation while keeping the elbow in contact with the towel to prevent compensatory movements [50].
		Side-Lying Forward Flexion	Lie on your side on a treatment table, holding a dumbbell with the shoulder in a neutral position. Perform 90° shoulder flexion in the sagittal plane [50].

Fig. 3.

Scapular Stabilizing Exercises (Strengthening)

Rotator cuff exercises

Regardless of the type of SD, all participants performed these exercises. Each exercise was performed for 3 sets of 10–12 repetitions. The exercises are illustrated in Fig. 4.

Fig. 4.

Rotator Cuff Exercises

Shoulder external and internal rotation

With the elbow flexed at 90° and kept close to the torso, one end of the resistance band is fixed while the participant holds the other end, keeping the forearm parallel to the ground. Externally rotate the forearm to activate the teres minor and infraspinatus. Internally rotate the forearm to engage the subscapularis.

Shoulder abduction

Secure one end of the resistance band under the foot while holding the other end with one hand, with the thumb pointing upward. Perform shoulder abduction along the scapular plane, lifting the resistance band to shoulder height to activate the deltoid and supraspinatus, then slowly lower it back down in a controlled manner.

Phase 3: additional functional shoulder exercises

After completing the initial two phases of training, participants’ shoulder stability was improved. The aim of this phase was to enhance multi-muscle coordination, dynamic joint stability, and neuromuscular control. These functional shoulder exercises replicate movement patterns encountered in daily life and sports, thereby promoting more effective comprehensive shoulder rehabilitation. The exercises are illustrated in Fig. 5.

Fig. 5.

Additional Functional Shoulder Exercises

Lawnmower exercise

Position: Secure one end of the resistance band to the floor. Hinge forward to approximately 45°, grasping the free end of the band with one hand while fully extending the arm. Place the other hand on the hip for stability.

Movement: Initiate the motion by pulling the resistance band diagonally upward and backward, mimicking the action of starting a lawnmower. Simultaneously, rotate the torso, retract the scapula toward the spine, and perform shoulder abduction and external rotation. Maintain a neutral spine throughout the movement.

Scapular push-ups

Position: Place your hands on the ground, shoulder width apart. Engage the core to maintain a straight alignment from the head to the feet, avoiding excessive lumbar extension or hip elevation.

Movement: Perform scapular retraction (downward movement) by drawing the scapulae toward the spine, engaging the middle trapezius and serratus anterior. Then, execute scapular protraction (upward movement) by actively pushing the ground away, allowing the scapulae to move laterally, emphasizing serratus anterior activation. Keep the arms fully extended throughout the exercise. This is a closed kinetic chain movement to enhance dynamic scapular control and shoulder stability.

Plank

Movement: Support the body on the forearms and elbows, with the legs extended and feet together, toes touching the ground. Maintain a straight alignment from head to toes, engage the core to prevent lumbar sagging or excessive arching, and avoid breath-holding. Hold the position for 20–30 s. With both forearms in stable contact with the floor, this closed kinetic chain exercise enhances core and scapular stability under static load.

Resistance band internal and external rotation at 90° shoulder abduction

Position: Secure one end of the resistance band at chest height and hold the other end. Position the shoulder at 90° abduction and the elbow at 90° flexion, with the scapula slightly depressed to stabilize the shoulder.

Movement: Slowly rotate the forearm externally, engaging the infraspinatus and teres minor. Then, rotate the forearm internally, activating the subscapularis. Hold at the end range for 1–2 s while maintaining the shoulder at 90° abduction and the elbow at 90° flexion throughout the movement.

Supplementary content

Throughout the intervention, most exercises followed a protocol of 3 sets of 10–12 repetitions per session. Red resistance bands, providing moderate resistance—greater than yellow bands but less than green bands—were used consistently. Dumbbells weighing 4 lbs were also employed. This protocol was developed based on clinical experience and preliminary pilot studies, which supported its feasibility. High participant adherence was observed, with minimal evidence of muscular compensation.

MPT group intervention program

The MPT group received multimodal physical therapy, consisting of three sequential components: ultrasound therapy, manual therapy, and range of motion exercises. The intervention was administered three times per week for a total of eight weeks, with each session lasting 60 min. The specific treatment protocol is detailed in Table 4.

Table 4.

MPT group intervention program

Section	Specific Procedures	Time Allocation
Part 1: Ultrasound Therapy	Using Chattanooga Mobile Ultrasound in pulsed mode (20% duty cycle), 1 MHz frequency, intensity of 1.0–1.5 W/cm² based on participant tolerance, for 8 min per session.	3 sessions per week, for 8 weeks
Part 2: Manual Therapy	1.Soft Tissue Mobilization Friction massage, myofascial release 2.Joint Mobilization Glenohumeral joint, acromioclavicular joint, scapula 3.Stretching Techniques Shoulder flexion, extension, abduction, orizontal abduction, external rotation, internal rotation	3 sessions per week, for 8 weeks
Part 3: Range of Motion Exercises	1.Wall Crawl Exercise 2.Pendulum Exercise 3.Stick Exercises	3 sessions per week, for 8 weeks

Part 1: ultrasound therapy for the shoulder

Ultrasound therapy has been recognized as an effective modality in physical medicine for treating acute and chronic musculoskeletal disorders [50]. It utilizes high-frequency sound waves to penetrate human tissues, producing deep thermal effects and micro-massage effects. These mechanisms promote soft tissue healing and pain relief by increasing tissue temperature [51], enhancing blood circulation [52], accelerating tissue regeneration [53], and improving tendon extensibility [54].

Previous studies have demonstrated that 15 sessions of 8-minute ultrasound therapy effectively alleviated shoulder pain in patients with subacromial impingement syndrome (SIS) and improved shoulder function and daily activity performance [55].

In this study, a Chattanooga mobile ultrasound device was used in pulsed mode (duty cycle: 20%) with a frequency of 1 MHz. The intensity was set between 1.0 and 1.5 W/cm², adjusted based on each participant’s subjective tolerance. The ultrasound was applied to a treatment area of approximately 25–50 cm², targeting the subacromial region. Each treatment session lasted 8 min, with an estimated total acoustic energy delivery ranging from 2,400 to 7,200 Joules per session. The intervention was administered three times per week over a period of 8 weeks, totaling 24 sessions.

Part 2: manual therapy

Based on the therapeutic interventions proposed in Therapeutic Exercise: Foundations and Techniques [56], the manual therapy component consists of three sequential techniques: soft tissue mobilization, joint mobilization, and stretching techniques.

Soft tissue mobilization

Friction massage

Applied to the supraspinatus tendon and the long head of the biceps tendon. The therapist uses the thumb to apply gentle pressure perpendicular to the muscle fibers, performing repeated friction movements. The goal is to reduce inflammation, enhance blood flow, and alleviate pain in the targeted musculotendinous structures.

Myofascial release

Applied to hypertonic muscle groups, including the upper trapezius, levator scapulae, pectoralis minor, and teres major. The pressure exerted on the muscles is gradually increased based on the participant’s tolerance. The objective is to release muscle tension, deactivate trigger points, and improve local circulation [57].

Joint mobilization

The intervention was primarily based on the Maitland concept of passive joint mobilization, targeting the acromioclavicular joint, glenohumeral joint, and scapula.

Glenohumeral joint distraction

The participant is positioned supine, with the arm in a resting position. The therapist stabilizes the humerus with both hands and applies a perpendicular force to the glenoid fossa, facilitating lateral movement of the humeral head away from the glenoid surface [58].

Inferior glide of the glenohumeral joint

The participant is positioned supine, with the arm abducted to 90°. The therapist stabilizes the humerus with both hands, applies a gentle distraction force, and directs the humeral head inferiorly along the glenoid fossa [59].

Superior glide of the glenohumeral joint

The participant is positioned seated, with the arm abducted to a comfortable range. The therapist stabilizes the humerus with both hands, applies a slight distraction force along the humeral axis, and glides the humeral head superiorly relative to the glenoid fossa [56].

Posterior glide of the glenohumeral joint

The participant is positioned supine, with the shoulder flexed to 90° and internally rotated, and the elbow flexed. One hand of the therapist applies a perpendicular distraction force to the glenoid fossa, while the other hand is placed at the participant’s elbow, directing it downward along the humeral axis to induce a posterior glide of the humeral head [60].

Anterior glide of the glenohumeral joint

The participant is positioned prone. The therapist stabilizes the humerus with both hands, applies a perpendicular distraction force, and directs the humeral head anteriorly by pressing downward at the posterior aspect of the shoulder [61].

Acromioclavicular joint glide

The participant is positioned seated. The therapist stabilizes the acromion with the lateral hand, while the medial hand is placed behind the clavicle near the joint space. The therapist then applies a forward-directed force to the clavicle using the thumb [56].

Scapular mobilization

The therapist mobilizes the scapula in elevation, depression, protraction, retraction, upward rotation, and downward rotation [62].

Stretching techniques

Passive stretching techniques are applied, in which the therapist passively moves the participant’s shoulder joint to its maximum ROM [56]. The stretching techniques include: shoulder flexion and extension, shoulder abduction, horizontal abduction, shoulder external rotation and internal rotation.

Part 3: range of motion exercise

This phase primarily aims to improve ROM and consists of three types of exercises. Each exercise is performed for 3 sets of 10–12 repetitions.

Wall crawl exercise

Position: The participant stands facing a wall with feet close to it, hands placed at shoulder width. The shoulders are slightly flexed, allowing the fingertips to just touch the wall.

Movement: The participant “crawls” their fingers up the wall by actively flexing the shoulder in the sagittal plane, reaching as high as possible within a tolerable range, before lowering the arm back down. This exercise promotes scapular anterior tilt and posterior depression in the sagittal plane.

Table-supported pendulum exercise

Position: The participant stands with feet shoulder-width apart, slightly leaning forward. The unaffected arm is placed on a table for support, while the affected arm hangs relaxed and fully extended.

Movement: The participant first performs gentle anterior-posterior and medial-lateral swinging motions with the relaxed arm. As mobility improves, small circular motions are introduced in a clockwise direction, followed by counterclockwise movements. The motion starts with a small range and gradually increases as flexibility improves [63].

Stick exercises

Position: The participant stands upright, holding a stick with both hands.

Movement: The participant actively performs the full ROM of the shoulder using a stick, including shoulder flexion, abduction and adduction, external and internal rotation, and horizontal abduction and adduction [56].

Intervention fidelity and participant adherence

Participant adherence in both the SDBET and MPT groups was tracked using attendance records, with adherence defined as attending at least 85% of sessions. Therapist fidelity was maintained using standardized checklists to document key intervention components, and verified through periodic audits by the lead investigator.

Variables of the study

This study examined both primary and secondary outcome variables. The primary variable was disability, measured using the Shoulder Pain and Disability Index (SPADI). The secondary variables included: (1) pain, assessed using the Visual Analog Scale (VAS) during the Hawkins-Kennedy Test; (2) shoulder active range of motion (AROM), measured using goniometry; (3) strength, evaluated by isometric strength tests (IST) targeting scapular stabilizers and external rotators; and (4) scapular kinematics, assessed using the Scapular Dyskinesis Test (SDT). All outcome measurements were conducted by two trained researchers at baseline, week 8 (post-intervention), and week 12 (4-week follow-up). Standardized protocols were strictly followed to minimize measurement variability and avoid unplanned analytical deviations.

Safety monitoring was conducted in accordance with CONSORT-Harms guidelines. Adverse events (AEs) were defined as any undesirable physical or psychological symptoms associated with the intervention, such as increased pain, dizziness, or muscle strain. Participants were instructed to immediately report any AEs to the research team. All reported AEs were documented using standardized adverse eve and no serious adverse events occurred during the study.

Primary variable

Disability

The Chinese version of the Shoulder Pain and Disability Index (SPADI) was used to quantify shoulder pain and disability [64]. SPADI is a validated questionnaire and is widely utilized in clinical practice for evaluating subacromial impingement syndrome (SIS). It is a self-administered questionnaire that measures the impact of shoulder problems over the past week. SPADI consists of two dimensions: pain and disability. The pain subscale includes five items evaluating the severity of shoulder pain, while the disability subscale comprises eight items assessing the difficulty experienced in performing various daily activities requiring upper limb use [65].

The internal consistency of the total score, as well as the pain and disability subscales, is high. The Cronbach’s alpha (α) coefficient is 0.9507 for the total index, 0.8604 for the pain subscale, and 0.9321 for the disability subscale [65]. The Chinese version of SPADI demonstrated Cronbach’s α coefficients ranging from 0.78 to 0.88, similar to those reported for the original version, indicating high internal consistency [64]. Furthermore, the intraclass correlation coefficient (ICC) for test-retest reliability was 0.874, suggesting good reliability [64].

The specific calculation formulas for the SPADI index are as follows:

Total Pain Score: 00 = %

(Note: If a respondent skips any item(s), the total possible score should be adjusted accordingly. For example, if one item is missing, the sum should be divided by 40 instead of 50.)

Total Disability Score: 100 = %

(Note: If a respondent skips any item(s), the total possible score should be adjusted accordingly. For example, if one item is missing, the sum should be divided by 70 instead of 80.)

Total SPADI Score: 100 = %

(Note: If a respondent skips any item(s), the total possible score should be adjusted accordingly. For example, if one item is missing, the sum should be divided by 120 instead of 130.)

Secondary variables

Pain

The Hawkins-Kennedy test has a sensitivity of 0.79 (95% CI: 0.75–0.82), indicating that 79% of patients with subacromial impingement syndrome (SIS) present a positive result. Its specificity is 0.59 (95% CI: 0.53–0.64) [66]. These findings demonstrate that the Hawkins-Kennedy test is highly sensitive in eliciting pain, making it a valuable tool for clinicians in assessing pain severity associated with SIS.

In this study, a 100 mm visual analog scale (VAS) was used to assess participants’ pain intensity during the Hawkins-Kennedy test. On this scale, 0 represents no pain, while 100 indicates the most severe pain imaginable [67].

Shoulder active range of motion

SIS commonly induces pain during shoulder abduction and flexion, as these movements increase contact or friction between the rotator cuff tendons and the acromial region. Therefore, this study measures the pain-free active range of motion (AROM) of the affected shoulder.

An electronic goniometer was used to assess active shoulder flexion and abduction AROM. The intra-class correlation coefficient (ICC-3,k) ≥ 0.94 indicates excellent intra-rater reliability for goniometric measurements [68]. The measurement procedure follows the guidelines established by Clarkson, which have been previously reported in the literature with high intra-rater reliability (ICC ≥ 0.85) [69].

AROM of shoulder flexion

The flexion AROM assessment was conducted with participants seated on a high-backed chair, ensuring that their lumbar spine remained in contact with the backrest to limit trunk compensation. Participants actively elevated their arm strictly within the sagittal plane, with the palm facing downward, until reaching their maximal active range. At this point, the measurement was recorded.

For inclinometric measurement, the instrument was placed proximally at the elbow and distally on the upper arm near the shoulder joint. For goniometric measurement, the fulcrum was positioned inferior and lateral to the acromion, the stationary arm was aligned parallel to the trunk, and the moving arm was aligned parallel to the longitudinal axis of the humerus.

AROM of shoulder abduction

The abduction AROM was measured in a seated position, with the trunk upright, similar to the flexion measurement. The arm was actively elevated strictly within the coronal plane, with the thumb pointing toward the ceiling to allow for the necessary external rotation, preventing greater tuberosity impingement against the acromion [70]. Once the maximal active range was reached, the measurement was recorded.

For inclinometric measurement, the instrument was placed proximally at the elbow and distally on the upper arm near the shoulder joint. For goniometric measurement, the fulcrum was positioned at the midpoint of the posterior aspect of the shoulder joint, the stationary arm was aligned parallel to the trunk, and the moving arm was aligned parallel to the longitudinal axis of the humerus.

Strength

Enhancing the strength of the middle trapezius (MT), lower trapezius (LT), and serratus anterior (SA) contributes to the coordinated recovery of scapular motion and stability. Additionally, shoulder external rotation (ER) at 90° of abduction is functionally significant in replicating real-life shoulder movements, particularly in assessing the function of rotator cuff muscles such as the supraspinatus, infraspinatus, and teres minor. Therefore, this study focuses on MT, LT, SA, and ER as key muscle strength assessment targets.

Isometric strength test (IST) was conducted using the Hoggan MicroFET2™ handheld dynamometer (HHD) to quantify muscle strength. Previous studies have employed HHD for IST of the LT, MT, and SA, demonstrating excellent test-retest reliability (ICC range: 0.89–0.96) [71]. The interrater reliability for shoulder external rotation strength measurement was ρ = 0.82, indicating good to excellent reliability, while the intrarater reliability was ρ = 0.92, reflecting excellent reliability [72].

The principle of IST is to apply resistance in the opposite direction of the targeted muscle contraction [71]. During measurement, the primary mover muscle should exhibit the highest level of activation, compared to secondary movers or stabilizers. The HHD is positioned at the midline of the specific muscle being tested, and resistance is gradually applied until the force exerted by the examiner equals that of the participant, achieving maximum isometric contraction. The peak isometric strength of the target muscle is then recorded [71]. If force equilibrium is not reached or the participant experiences pain, the test is halted and resumed after a brief rest period. A 2-minute rest was provided between each muscle strength test. The dominant side muscle strength was tested three times, and the average of the three highest values (measured in kg) was used for data analysis [73].

IST of MT

Scapular retraction is the primary movement when the MT contracts. The procedure follows the description by Hislop et al. [74]. The HHD resistance was applied on the scapular spine. Force was applied horizontally, parallel to the long axis of the humerus, with the humerus abducted at 90°.

IST of LT

Scapular retraction and depression occur when the LT contracts. The procedure follows Hislop et al. [74]. The HHD resistance was applied to the scapular spine. Force was applied superolaterally along the scapular spine, parallel to the long axis of the humerus, with the humerus positioned at an elevation angle of 140°.

IST of SA

Scapular protraction is the primary movement when the SA contracts. The procedure follows Kendall et al. [75]. The participant flexes elbow at 90°, positions the shoulder in 90° of flexion and internal rotation, and tester apply HHD resistance along the long axis of the humerus at the olecranon of the ulna.

IST of ER

The ER strength test was conducted in the supine position to minimize the influence of limb weight differences. During the test, the shoulder was abducted to 90°, and the elbow was flexed to 90°. The HHD resistance was applied perpendicularly to the posterior forearm, approximately one inch proximal to the ulnar styloid process. The participant performs shoulder external rotation, while the examiner applies opposing force using HHD. The participant is instructed to reach maximum tension within 1–2 s and maintain this tension for 4–5 s [72, 76].

Scapular kinematics

In this study, the Scapular Dyskinesis Test (SDT) was conducted using Kibler’s assessment method [41]. The evaluator observed participants during bilateral humeral abduction and elevation movements and categorized their scapular movement patterns into four types: Type I is characterized by prominence of the inferomedial scapular angle; Type II presents as prominence of the entire medial scapular border; Type III is defined by prominence of the superior scapular border [41, 77]; Type IV is considered symmetrical scapular motion and is defined as a normal scapular movement pattern.

All participants were filmed from a posterior view in a standing, relaxed posture. A camera mounted on a table was positioned 204 cm away on a fixed stand. Participants were instructed to elevate their arms in the frontal plane (abduction) and at 45° anterior to the frontal plane. Adequate lighting was provided to ensure clear visualization of scapular motion in the video recordings. Each participant performed three repetitions of bilateral arm elevation and abduction in a balanced sequence to prevent fatigue [41].

Each participant’s videos and images were coded for identification and subsequently reviewed by blinded evaluators. The recorded footage was assessed by two licensed physical therapists and two kinesiologists, all of whom had received specific training for this study. These evaluators were not involved in data collection to minimize bias. The inter-rater reliability of this method was 0.4, while the intra-rater reliability was 0.5, indicating moderate consistency and reliability [41].

Data analysis

All data were analyzed using SPSS 27.0 (IBM, Armonk, NY, USA). No missing data were observed during the study period. Therefore, no imputation techniques were required. All analyses were conducted according to the intention-to-treat (ITT) principle, including all randomized participants in their originally assigned groups.

For all outcome measures except SDT (including SPADI, VAS, AROM, and IST), mixed-design ANOVA was employed. Time (T0, T8, T12) was considered the within-subject factor, and group (SDBET vs. MPT) was considered the between-subject factor. When a significant interaction effect was observed, simple effects analysis was performed to further explore specific differences between time points and groups.

Simple Effects Analysis: Between-group comparisons: A linear mixed model was used to compare group differences at each time point, with Bonferroni correction applied for multiple comparisons. Within-group comparisons: Repeated measures ANOVA was used to compare differences within each group across time points, also with Bonferroni correction for multiple comparisons.

Before conducting statistical analyses, all data (except for SDT) underwent the Shapiro-Wilk test to assess normality. Data that met the normality assumption were reported as mean ± standard deviation (Mean ± SD). All analysis results reported 95% confidence intervals and effect sizes (e.g., partial η²). For SDT, due to the small frequency of certain categories (i.e., frequency < 5 in some types), Fisher’s exact test was used for between-group comparisons to ensure more accurate statistical results.

Result

A total of 68 overhead athletes were screened for eligibility. Of these, 36 were excluded due to not meeting the inclusion criteria (n = 23), declining to participate (n = 6), or other reasons (n = 7). The remaining 32 participants were enrolled and randomly assigned to either the SDBET group (n = 16) or the MPT group (n = 16). All participants completed the full 8-week intervention and 4-week follow-up with full adherence. No dropouts or missing data were reported during the study period. Participants (aged 18–25 years) were recruited from a local sports academy (n = 13), an athletic club (n = 8), and a rehabilitation department (n = 11). No serious adverse events occurred. However, two participants in the SDBET group experienced mild shoulder impingement pain during Phase 1, likely due to insufficient scapular upward rotation and improper movement technique. These symptoms were resolved through targeted movement correction, and both participants completed the intervention without further complications.

There were no significant baseline differences between the SDBET and MPT groups across all measured variables, including demographics, disability, pain, range of motion, strength, and scapular kinematics (all p > 0.05). Baseline demographic data are shown in Table 5.

Table 5.

Participant demographics: age, height, and weight (Mean ± SD)

Variable	SDBET group (n = 16)	MPT group (n = 16)	T	P
Age (years)	21.06 ± 2.52	20.63 ± 2.34	0.51	0.614
Height (cm)	174.63 ± 5.50	174.19 ± 3.75	0.26	0.794
Weight (kg)	67.85 ± 4.61	68.61 ± 5.52	-0.42	0.677

Mixed-design ANOVA results

The results of the mixed-design ANOVA are presented in Table 6. Significant main effects of time were observed for each outcome variable (all p < 0.001), with large effect sizes (partial η² ranging from 0.71 to 0.92). Significant main effects of group were found for SPADI (F = 8.52, p = 0.007, η² = 0.22), MT (F = 46.91, p < 0.001, η² = 0.61), LT (F = 52.92, p < 0.001, η² = 0.64), SA (F = 28.41, p < 0.001, η² = 0.49), and SER (F = 22.84, p < 0.001, η² = 0.43). Moreover, significant Time × Group interaction effects were identified for each outcome variable (all p < 0.05), with effect sizes ranging from η² = 0.15 to 0.76. Given the presence of significant interaction effects, further simple effects analyses were conducted to explore between-group differences at each time point and within-group changes over time. The results are detailed in Table 6.

Table 6.

Results of mixed-design ANOVA for each outcome measure

Outcome Measure	Time			Group			Interaction
	F	P	Partial η²	F	P	Partial η²		F	P	Partial η²
SPADI	335.17	< 0.001	0.92	8.52	0.007	0.22		5.19	0.023	0.15
VAS	220.90	< 0.001	0.88	0.16	0.689	0.01		8.70	< 0.001	0.23
Shoulder Flexion AROM	120.53	< 0.001	0.80	2.57	0.120	0.08		14.37	< 0.001	0.32
Shoulder Abduction AROM	137.91	< 0.001	0.82	2.74	0.109	0.08		13.86	< 0.001	0.32
MT (kg)	100.79	< 0.001	0.77	46.91	< 0.001	0.61		92.70	< 0.001	0.76
LT (kg)	74.37	< 0.001	0.71	52.92	< 0.001	0.64		65.13	< 0.001	0.69
SA (kg)	99.04	< 0.001	0.77	28.41	< 0.001	0.49		65.80	< 0.001	0.69
SER (kg)	81.87	< 0.001	0.73	22.84	< 0.001	0.43		67.88	< 0.001	0.69

Mixed-design ANOVA was used to analyze the effects of Time (within-subjects factor: T0, T8, T12), Group (between-subjects factor: SDBET vs. MPT), and their interaction on each outcome measure. Partial η² indicates effect size

Between- and within-group comparisons of outcome measures

SPADI

Between-group comparisons showed no significant difference in SPADI scores between the SDBET and MPT groups at week 8 (p = 0.316, η² = 0.03). However, at week 12, the SDBET group demonstrated significantly lower SPADI scores compared to the MPT group (p < 0.001, 95% CI [− 7.85, − 4.12]), with a large effect size (η² = 0.59), indicating a meaningful between-group difference favoring the SDBET group.

Within-group comparisons revealed significant improvements in SPADI from baseline to week 8 in both the SDBET and MPT groups (p < 0.05), with mean changes exceeding the established minimal clinically important difference (MCID) range of 14.1 to 20.6 points [40]. However, from week 8 to week 12, SPADI scores in the MPT group significantly increased (p < 0.05), suggesting a decline in treatment effect, while the SDBET group maintained stable scores with no significant change (p > 0.05). Results are presented in Tables 7 and 8.

Table 7.

Descriptive statistics and Between-Group comparisons

Outcome Measures	Time	SDBET group (Mean ± SD)	MPT group (Mean ± SD)	Partial η²	95%CI	P
SPADI	Baseline	45.67 ± 6.01	46.46 ± 5.34	0.01	(−4.89, 3.32)	0.698
	Week 8	24.20 ± 2.55	25.12 ± 2.58	0.03	(−2.78, 0.93)	0.316
	Week 12	23.37 ± 3.17	29.35 ± 1.80	0.59	(−7.85,−4.12)	< 0.001*
VAS	Baseline	45.38 ± 5.40	45.50 ± 5.69	0.01	(−4.13, 3.88)	0.950
	Week 8	30.94 ± 5.79	26.38 ± 4.40	0.17	(0.85, 8.27)	0.018*
	Week 12	30.63 ± 6.85	33.00 ± 4.90	0.04	(−6.68, 1.93)	0.268
Shoulder Flexion AROM (degrees)	Baseline	159.38 ± 6.50	157.94 ± 6.02	0.01	(−3.09, 5.96)	0.521
	Week 8	169.88 ± 5.32	170.38 ± 4.99	0.01	(−4.22, 3.22)	0.786
	Week 12	171.13 ± 5.15	163.69 ± 5.07	0.37	(3.75, 11.13)	< 0.001*
Shoulder Abduction AROM (degrees)	Baseline	146.06 ± 7.78	143.75 ± 8.39	0.02	(−3.53, 8.16)	0.425
	Week 8	161.44 ± 5.28	163.19 ± 5.78	0.03	(−5.75, 2.25)	0.378
	Week 12	162.63 ± 5.43	153.06 ± 7.35	0.37	(4.90, 14.23)	< 0.001*
MT (kg)	Baseline	10.23 ± 0.71	9.97 ± 0.64	0.04	(−0.23, 0.74)	0.295
	Week 8	12.20 ± 0.70	10.00 ± 0.67	0.74	(1.71, 2.69)	< 0.001*
	Week 12	12.40 ± 0.82	10.03 ± 0.79	0.70	(1.80, 2.96)	< 0.001*
LT (kg)	Baseline	7.89 ± 0.62	7.70 ± 0.63	0.02	(−0.26, 0.64)	0.399
	Week 8	9.87 ± 0.71	7.80 ± 0.57	0.74	(1.62, 2.54)	< 0.001*
	Week 12	9.96 ± 0.82	7.75 ± 0.62	0.71	(1.69, 2.74)	< 0.001*
SA (kg)	Baseline	14.96 ± 0.71	14.83 ± 0.79	0.01	(−0.41, 0.68)	0.619
	Week 8	17.06 ± 0.84	15.01 ± 0.83	0.62	(1.45, 2.65)	< 0.001*
	Week 12	17.17 ± 0.87	15.08 ± 0.87	0.61	(1.46, 2.71)	< 0.001*
SER (kg)	Baseline	6.81 ± 0.93	6.77 ± 1.08	0.01	(−0.69, 0.77)	0.912
	Week 8	9.14 ± 0.92	6.91 ± 1.06	0.57	(1.51, 2.95)	< 0.001*
	Week 12	9.24 ± 0.83	6.85 ± 1.08	0.62	(1.69, 3.09)	< 0.001*

* Indicates p < 0.05 for between-group comparisons at the same time

Table 8.

Within-Group comparisons for the SDBET and MPT groups across time points

Outcome Measure	Time Comparison	SDBET Group		MPT Group
		CI	P	CI	P
SPADI	Baseline vs. Week 8	(16.89, 26.05)	< 0.001*	(17.84, 24.83)	< 0.001*
	Baseline vs. Week 12	(17.37, 27.23)	< 0.001*	(13.77, 20.44)	< 0.001*
	Week 8 vs. Week 12	(-1.14, 2.80)	0.823	(-5.24, -3.22)	< 0.001*
VAS	Baseline vs. Week 8	(10.57, 18.30)	< 0.001*	(15.82, 22.43)	< 0.001*
	Baseline vs. Week 12	(10.61, 18.89)	< 0.001*	(9.96, 15.04)	< 0.001*
	Week 8 vs. Week 12	(-2.11, 2.73)	> 0.999	(-9.34, -3.91)	< 0.001*
Shoulder Flexion AROM	Baseline vs. Week 8	(-13.66, -7.34)	< 0.001*	(-16.22, -8.65)	< 0.001*
	Baseline vs. Week 12	(-14.89, -8.62)	< 0.001*	(-8.34, -3.17)	< 0.001*
	Week 8 vs. Week 12	(-3.44, 0.94)	0.436	(4.18, 9.19)	< 0.001*
Shoulder Abduction AROM	Baseline vs. Week 8	(-18.60, -12.15)	< 0.001*	(-24.83, -14.05)	< 0.001*
	Baseline vs. Week 12	(-21.14, -11.98)	< 0.001*	(-13.98, -4.65)	< 0.001*
	Week 8 vs. Week 12	(-4.03, 1.65)	0.832	(6.52, 13.73)	< 0.001*
MT (kg)	Baseline vs. Week 8	(-2.42, -1.53)	< 0.001*	(-0.17, 0.11)	> 0.999
	Baseline vs. Week 12	(-2.66, -1.69)	< 0.001*	(-0.28, 0.17)	> 0.999
	Week 8 vs. Week 12	(-0.51, 0.11)	0.291	(-0.25, 0.20)	> 0.999
LT (kg)	Baseline vs. Week 8	(-2.57, -1.40)	< 0.001*	(-0.22, 0.03)	0.204
	Baseline vs. Week 12	(-2.71, -1.43)	< 0.001*	(-0.20, 0.11)	> 0.999
	Week 8 vs. Week 12	(-0.33, 0.15)	0.995	(-0.06, 0.15)	0.859
SA (kg)	Baseline vs. Week 8	(-2.53, -1.67)	< 0.001*	(-0.43, 0.07)	0.209
	Baseline vs. Week 12	(-2.68, -1.73)	< 0.001*	(-0.62, 0.11)	0.236
	Week 8 vs. Week 12	(-0.41, 0.20)	> 0.999	(-0.44, 0.28)	> 0.999
SER (kg)	Baseline vs. Week 8	(-3.04, -1.63)	< 0.001*	(-0.31, 0.02)	0.100
	Baseline vs. Week 12	(-3.14, -1.73)	< 0.001*	(-0.25, 0.08)	0.591
	Week 8 vs. Week 12	(-0.30, 0.10)	0.617	(-0.10, 0.23)	0.981

*Indicates p < 0.05 for within-group comparisons at different time

VAS

Between-group comparisons indicated that at week 8, the reduction in VAS scores in the MPT group was significantly greater than that in the SDBET group (p = 0.018, 95% CI [0.85, 8.27]), with a moderate effect size (η² = 0.17). However, by week 12, VAS scores in the MPT group increased, and no significant difference was observed between the two groups (p = 0.268, η² = 0.04).

Within-group comparisons demonstrated significant reductions in VAS scores from baseline to week 8 in both the SDBET and MPT groups (p < 0.05), with mean reductions exceeding the minimal clinically important difference (MCID) threshold of 14 mm [78]. However, following intervention cessation, the MPT group showed a significant rebound in VAS scores at week 12 (p < 0.05), whereas the SDBET group maintained stable scores without significant changes (p > 0.05). Results are presented in Tables 7 and 8.

AROM

Between-group comparisons showed no significant difference in AROM for shoulder flexion and abduction between the two groups at week 8 (p = 0.786 and 0.378, respectively; η² = 0.01 and 0.03). However, at week 12, the SDBET group showed significantly greater shoulder flexion (p < 0.001, 95% CI [3.75, 11.13], η² = 0.37) and abduction (p < 0.001, 95% CI [4.90, 14.23], η² = 0.37) AROM than the MPT group, indicating a moderate-to-large effect size favoring the SDBET intervention.

Within-group comparisons demonstrated significant improvements in both shoulder flexion and abduction AROM from week 0 to week 8 in both groups (p < 0.05). However, at week 12, AROM values in the MPT group declined compared to week 8 (p < 0.05), whereas the SDBET group maintained improvements with no significant changes (p > 0.05). These findings suggest that the improvements in AROM were more sustained in the SDBET group. Results are presented in Tables 7 and 8.

IST

Between-group comparisons revealed that at both week 8 and week 12, isometric strength of the middle trapezius (MT), lower trapezius (LT), serratus anterior (SA), and shoulder external rotators (SER) was significantly higher in the SDBET group compared to the MPT group (all p < 0.001), with large effect sizes observed at both time points. Specifically, at week 8, partial η² values were 0.74 for MT, 0.74 for LT, 0.62 for SA, and 0.57 for SER. At week 12, the corresponding partial η² values were 0.70, 0.71, 0.61, and 0.62, respectively.

Within-group comparisons showed that muscle strength in the MPT group did not significantly improve throughout the intervention period. In contrast, the SDBET group demonstrated significant and consistent improvements in all four muscle groups at both week 8 and week 12 (p < 0.05). These improvements were maintained even after the intervention ceased, suggesting that the SDBET program effectively and sustainably enhanced scapular stabilizing muscle strength. Results are presented in Tables 7 and 8.

SDT

Between-group comparisons revealed significant differences in SDT scores between the SDBET and MPT groups at both week 8 and week 12 (p < 0.05). The results are presented in Table 9.

Table 9.

Between-Group comparison of SDT

Time	Type of SD	SDBET Group (n = 16)	MPT Group (n = 16)	Fisher’s Exact Test	P
Baseline	Type I	7	5	0.72	0.894
	Type II	6	8
	Type III	3	3
	Type IV	0	0
Week 8	Type I	4	5	9.67	0.018*
	Type II	4	8
	Type III	1	3
	Type IV	7	0
Week 12	Type I	4	5	9.67	0.018*
	Type II	4	8
	Type III	1	3
	Type IV	7	0

Fisher’s exact test was used for between-group comparisons

P < 0.05 indicates statistical significance

Within-group comparisons showed that, by week 8, seven out of 16 participants (43.8%) in the SDBET group had improved their scapular movement pattern to Type IV, representing a significant improvement from baseline (p < 0.05). Moreover, at week 12, these improvements in the SDBET group remained stable, with no observed regression (p > 0.05), indicating the long-term effectiveness of SDBET in improving SD. In contrast, no significant improvement in SD was observed in the MPT group at either week 8 or week 12 (p > 0.05), and none of the participants returned to normal scapular movement (Type IV).

Discussion

This study aimed to compare the effectiveness and underlying mechanisms of scapular dyskinesis-based exercise therapy (SDBET) and multimodal physical therapy (MPT) in young overhead athletes with subacromial impingement syndrome (SIS) and scapular dyskinesis (SD). The results showed that both interventions led to significant improvements in disability, pain, and active range of motion by week 8. Notably, the MPT group exhibited greater pain relief than the SDBET group at week 8. However, the SDBET group demonstrated additional gains in scapular stabilizer strength, external rotator strength, and scapular kinematics. Moreover, the SDBET group maintained improvements in disability, pain, and active range of motion at the 12-week follow-up, whereas these outcomes regressed in the MPT group. These findings suggest that targeting scapular dyskinesis may offer more sustained clinical benefits in the management of SIS in this population.

Analysis of disability

Both SDBET and MPT interventions effectively improved SPADI. This aligns with Struyf et al. [79], who reported that a scapula-focused program—including stretching, scapular control, and rotator cuff strengthening—significantly improved shoulder pain and function in SIS patients. Similarly, Roy et al. [80] found that combining motor control and strengthening exercises led to significant improvements in SPADI scores and pain, consistent with the SDBET approach used in this study.

However, unlike previous studies, the current findings reveal a difference in the sustainability of treatment effects. At the 12-week follow-up, SPADI in the MPT group had regressed compared to week 8, whereas the SDBET group maintained its improvements. Firstly, the sustained benefits observed in the SDBET group may be attributed to the integrative intervention, which combined scapular control, scapular stabilization, rotator cuff and functional shoulder exercises. This approach likely enhanced proprioceptive input, neuromuscular control, and scapular muscle strength [80], contributing to the maintenance of improved shoulder function. Secondly, the phased and progressively loaded design of the SDBET program improved participants’ tolerance to moderate-to-high intensity shoulder activities. Finally, the inclusion of multiple closed kinetic chain (CKC) exercises across all three phases likely contributed to improved shoulder stability and proprioception [34, 81], enabling participants to maintain functional performance in daily and athletic tasks such as pushing, pulling, and lifting even after the intervention.

In contrast, the MPT group primarily relied on passive modalities, such as ultrasound and manual therapy. Although these methods provided short-term symptom relief, they did not fundamentally address deficits in scapular motion or shoulder joint stability. Consequently, upon resumption of regular physical activity, participants in the MPT group may have been re-exposed to biomechanical overload without adequate neuromuscular preparedness, resulting in symptom recurrence and functional decline.

Analysis of pain

Both intervention groups demonstrated significant improvements in VAS, indicating effective pain reduction. Notably, this study found that the MPT group exhibited superior pain relief compared to the SDBET group at week 8. This short-term analgesic effect is likely attributable to the multimodal nature of the MPT intervention, which integrated several therapeutic approaches that efficiently alleviated shoulder pain:

Ultrasound therapy, which promotes deep soft tissue repair, elevates pain thresholds, and reduces deep muscle tension [53, 82];

Manual therapy, including joint mobilizations, soft tissue techniques, and passive stretching, which relieve pain, reduce muscular tightness, and improve joint mobility [83];

Range of motion exercises, aimed at enhancing shoulder flexibility and function while minimizing joint impingement.

Given that the SDBET intervention primarily focused on improving muscular function and neuromuscular control, its initial pain-relieving effect may have been less pronounced compared to the MPT protocol, which was more directly targeted at symptom alleviation.

However, despite the superior short-term pain reduction in the MPT group, this effect was not sustained beyond the intervention period. By the 12-week follow-up, VAS scores had regressed in the MPT group. In contrast, the SDBET group, although demonstrating slightly slower pain relief initially, maintained stable improvements in VAS scores during follow-up. This long-term benefit may be attributed to the intervention’s emphasis on scapular control and shoulder stability, which likely addressed the underlying biomechanical contributors to impingement. By improving dynamic scapular motion and enhancing neuromuscular coordination, the SDBET group may have developed greater resilience against re-injury, thereby reducing the likelihood of pain recurrence after the cessation of treatment.

Analysis of shoulder active range of motion

In patients with SIS, soft tissue compression during shoulder flexion and abduction often induces pain, leading to restricted shoulder joint mobility. The results of this study indicate that SDBET and MPT exhibit comparable short-term efficacy in improving pain-free active range of motion (AROM) of the shoulder joint. This finding aligns with the study by Azar Moezy et al. [29].

In this study, although both SDBET and MPT significantly improved AROM, their intervention mechanisms differed. First, the SDBET group incorporated scapular control exercise, which enhanced proprioception and neuromuscular control, optimizing scapular upward rotation and shoulder elevation coordination. This resulted in reduced compression of the subacromial space and an increased AROM range [84]. Second, scapular stabilization exercise in the SDBET group strengthened scapular muscles, relaxed overactive muscles, and improved scapular stability and movement coordination. Finally, rotator cuff strengthening exercises in the SDBET group effectively stabilized the humeral head, preventing abnormal superior migration or displacement, thereby reducing impingement against the acromion. For the MPT group, ultrasound therapy reduced inflammation and accelerated tissue repair, significantly alleviating pain, with increased shoulder ROM likely attributed to pain relief [85]. Manual therapy and ROM exercises improved soft tissue extensibility and joint mobility, but lacked long-term neuromuscular adaptations.

However, unlike previous studies, this study found that at the 4-week follow-up after intervention cessation, pain-free shoulder flexion and abduction AROM regressed in the MPT group, whereas the SDBET group maintained its improvements. This suggests that while MPT effectively enhanced shoulder mobility in the short term through anti-inflammatory and analgesic effects, its benefits were not sustained due to the lack of targeted training for scapular dyskinesis—such as restoring scapular movement patterns, correcting muscle imbalances, and strengthening the rotator cuff.

Analysis of strength

In this study, improvements in IST in the SDBET group were accompanied by reductions in SIS and SD symptoms, suggesting that enhanced strength of the scapular stabilizers and rotator cuff muscles contributes to better shoulder function. These findings are consistent with those of Baskurt et al. [27], who implemented a six-week intervention involving stretching, strengthening, and scapular stabilization exercises in SIS patients, resulting in significant improvements in the strength of MT, LT, SA, and rotator cuff muscles, along with symptom relief. Similarly, Hotta et al. [86] reported that a scapula-focused exercise program combining scapular motor control and strengthening effectively improved MT, LT, and SA strength, as well as shoulder function and scapular kinematics.

Notably, the SDBET group maintained scapular muscle strength even after the intervention period, a finding not commonly reported in previous studies. This sustained improvement is likely attributable to increased neuromuscular activation and progressive strengthening strategies. The combination of scapular control exercises and closed kinetic chain (CKC) exercises may have enhanced muscle activation [87, 88], while concurrent resistance exercise further facilitated muscular adaptation. In contrast, the MPT group emphasized symptomatic relief without adequately improving muscle function. Consequently, residual issues such as muscle imbalance and insufficient muscle activation may persist, potentially increasing the risk of recurrent impingement during overhead activities post-intervention.

Analysis of scapular kinematics

In this study, 7 participants (43.8%) in the SDBET group demonstrated improvement of SD, representing a statistically significant change. This improvement can be attributed to three key factors. First, the incorporation of scapular control and stabilization exercises likely enhanced scapular motor control and stability, thereby correcting abnormal scapular kinematics [86]. Second, the scapular stabilization exercise in the SDBET group was adapted from the protocol developed by Lan Tang et al. [49] and further modified to address different types of SD, allowing for a more individualized approach. Tang et al. [85] reported that 31.71% of patients with scapular dyskinesis achieved normalized scapular motion following a targeted intervention, whereas no such improvement was observed in those who received only range of motion exercises. The present findings align with and further support the effectiveness of this targeted intervention strategy. Third, the inclusion of multiple closed kinetic chain (CKC) exercises in the SDBET program likely contributed to the restoration of scapular kinematics by promoting joint stability and neuromuscular control [89]. In contrast, the MPT intervention did not specifically target scapular control or stability, which may explain its limited long-term impact on scapular kinematics.

Study limitations

First, the sample size was relatively small, and all participants were male overhead athletes aged 18 to 25 years, which limits the generalisability of the findings. The results may not be applicable to females, older adults, or individuals with SIS who do not engage in overhead sports. Future larger, multicenter trials with more diverse populations are needed to confirm the clinical efficacy and improve the external validity of SDBET.

Second, the four-week follow-up period post-intervention may be insufficient to assess the long-term sustainability of treatment effects or the risk of symptom recurrence. Although short-term follow-up is common in preliminary trials to assess immediate treatment outcomes, future studies are warranted to include longer-term follow-ups (e.g., ≥ 6 months) to better evaluate relapse rates and long-term efficacy of the interventions.

Third, pain-related muscle inhibition during isometric strength testing may have affected the accuracy of force measurements. Incorporating pain management strategies or alternative assessment methods may help reduce this bias.

Fourth, this study did not assess scapular kinematics in detail. Future research should consider using advanced technologies, such as 3D motion analysis or surface electromyography, to better evaluate changes in scapular movement and explore biomechanical and neuromuscular mechanisms.

Fifth, participants and therapists could not be blinded due to the nature of the intervention, which may have introduced performance bias. Although outcome assessors were blinded, the absence of a blinding-success check may pose a risk of detection bias. Standardized protocols and objective measures were used to minimize this risk.

Finally, the combined effect of SDBET and MPT was not investigated. Future studies could examine whether their integration offers faster symptom relief and greater long-term benefits.

Conclusion

This study evaluated the efficacy of SDBET and MPT in SIS patients with SD. Both SDBET and MPT significantly improved patients’ shoulder disability, shoulder ROM, and pain relief. However, compared with SDBET, MPT lacked sustained therapeutic effects. SDBET showed more lasting efficacy in SIS patients with SD by effectively improving SD, restoring scapular control, enhancing scapular muscle activation, releasing tight muscles, and improving scapular and rotator cuff muscle strength. Therefore, the implementation of targeted treatment may be the future development direction of physical therapy in order to obtain more comprehensive and long-term therapeutic effects.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

ML W and XMH jointly contributed to the writing of the manuscript. Specifically, XMH was responsible for the entire intervention process, study design, and manuscript drafting; MLW was in charge of data analysis and manuscript drafting; GYB was responsible for randomization, oversight activities, and resource provision.

Funding

This study received no financial support from any institute or agency. However, [the Rehabilitation Department of Tongliao Mongolian Medicine Hospital] provided ultrasound therapy equipment and access to laboratory facilities during the study.

Data availability

Availability of data and material: The researchers interested in using the final dataset for scientific purposes may contact the corresponding author.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Tongliao Mongolian Medical Hospital (Approval No: [TLSMYYY-2024-3-002]) and informed consent was obtained from all individual participants included in the study. This study performed on human data or materials complied with the Declaration of Helsinki.

Consent for publication

Written informed consent was obtained from all study participants for the publication of their clinical information. The individual depicted in the published image is one of the authors and has provided explicit written consent for the publication of this identifying image.

Competing interests

The authors declare no competing interests.