Introduction

Rotator Cuff-Related Shoulder Pain (RCRSP) is a condition in which pain leads to decreased shoulder functioning. This can greatly impact individuals’ daily functioning, leading to seeking care. Exercise is regarded as the mainstay of rehabilitation but no consensus exists on the type of exercise that is preferred. Recently, a consensus decision algorithm was published by shoulder experts to guide clinical reasoning and decision-making. The algorithm starts from clinical findings rather than from the structural pathology to guide reasoning and treatment. This algorithm has been studied extensively but the clinical value could not be supported or refuted. The evidence quality of the systematic review was however downgraded by inadequate sample sizes, high risk of selection bias, and poor descriptions of the interventions. This randomized controlled trial ought to conduct a methodologically high-quality study to study the effectiveness of the proposed exercises for RCRSP.

Methods

This study was a single-blind, randomized controlled trial with two parallel groups. The participants were recruited from the Physiotherapy Department of the Clinical Hospital San Borja Arriaran. They had a diagnosis of RCRSP based on the WHO International Classification of Diseases-10 criteria. This includes:

• Compression of the rotator cuff tendons and subacromial bursa between the humeral head and structures that make up the coracoacromial arch and the humeral tuberosities.
• This condition is associated with subacromial bursitis and rotator cuff (largely supraspinatus) and bicipital tendon inflammation, with or without degenerative changes in the tendon.
• Pain that is most severe when the arm is abducted in an arc between 40 and 120 degrees, sometimes associated with tears in the rotator cuff, is the chief symptom

Eligible patients were adults aged 18 or older with shoulder symptoms for a minimum of three months. At least three positive clinical signs from the following tests were required:

• Neer Test,
• Hawkins-Kennedy test,
• A painful arc,
• Pain on resisted external rotation
• Exclusion criteria: Individuals with cervical radiculopathy, arthritis in shoulder joints, rotator cuff tears, prior shoulder surgery, or corticosteroid injection in the last 12 months were excluded.

Eligible participants were randomly assigned to the experimental or control groups. The intervention group received the specific exercises for RCRSP while the control group performed the general exercises. All participants continued to receive their usual care which consisted of taking 2 times per day 500mg of oral naproxen for 14 days and engaging in a supervised home exercise program. Every participant regardless of group allocation was given a session of advice in which they also learned how to perform the home exercises. Four exercises were given in the home exercise program and had to be pain-free. They had to be performed 10 times each, 2 times per day. Every week, the participants reviewed the exercises with the physiotherapist. The home exercise program included:

• Shoulder elevation
• Shoulder retraction
• Shoulder abduction in the scapular plane
• Neck retraction

Experimental group

In addition to the usual care, as specified above, the participants who were randomly assigned to the intervention group were given a supervised specific exercise program based on the clinical decision algorithm. This program lasted 5 weeks and had to be completed 2 times per week. The program included:

• Scapular Orientation: Exercises aimed at improving scapular positioning and proprioception.
• Scapular Control Exercises: Bilateral shoulder flexion up to 60°, unilateral bench press, and prone shoulder retraction and extension.
• Glenohumeral Control Exercises: Isometric external rotation with shoulder adduction, isometric adduction of the shoulder in the scapular plane at 30° and 60° of elevation

Exercises were progressed based on participant tolerance, with a focus on pain-free movement, specific muscle activation, and gradual progression in difficulty. Each session included 8–10 repetitions per exercise, with a 5–10 second hold and 30 seconds to 1-minute rest between repetitions.

Control Group 

This group performed a general exercise program, supervised twice a week for five weeks. The general exercises included exercises for overall shoulder strengthening, stretching, and mobility but lacked the specificity of targeted muscle activation and coordination emphasized in the specific exercise group.

The “core programme” as can be seen in the table below was carried out over the first three to four treatment sessions. Patients did dynamic exercises twice a week, beginning with two sets of 10 repetitions and low resistance (yellow rubber band). Shoulder and neck stretches were held for 10 seconds and performed twice. The isometric scapular training positions were held for 10 seconds and repeated twice.

Progressions were made if patients completed the main program without difficulties. Sets were raised from two to three. The repetitions (or seconds for static exercises) were raised from 10 to 20. In the final step, resistance was raised from yellow to red and green rubber bands.

Exercises from a “additional program” were added if the patient could still complete the core program without difficulty. For example, exercise C3 was replaced by A4, C4 by A5, and C6 by A6.

Patients were required to cease exercising if they experienced discomfort greater than 3 out of 10 on a VAS or for more than 30 seconds after stopping the exercise. The physiotherapists adjusted exercise resistance, sets, repetitions, and range of motion to reduce discomfort. Alternative versions of several exercises could be employed, such as C6b instead of C6a. If an activity was unable to be performed due to pain, it was substituted by AP1 and AP2 in the following two sessions.

Outcome Measures

Primary Outcome: Shoulder function was measured using the Shoulder Pain and Disability Index (SPADI). The minimal clinically important difference is 20 points.Secondary Outcomes: These included the Disabilities of the Arm, Shoulder, and Hand (DASH) Questionnaire for upper limb function, the Visual Analog Scale (VAS) for pain intensity, and the Tampa Scale of Kinesiophobia (TSK) for fear of movement.

Results

52 participants were recruited and equally divided into the experimental and control groups. There were no significant differences in their baseline characteristics.

Shoulder function improved in both groups as can be seen in the raincloud plots below. The experimental group performing specific shoulder exercises improved to a greater extent compared to the control group. The mean between-group difference at 5 weeks was 13.5 points, favoring the experimental group. 

The secondary outcomes supported the primary outcome:

• Upper limb function improved in both groups but to a larger extent in the experimental group, leading to a mean between-group difference of 10.1 points.

• Pain intensity at rest decreased slightly in both groups but the mean between-group difference was only 0.2cm. Hence, none of the interventions was superior regarding pain decrease at rest. 

• Pain intensity during movement decreased in both groups but with a larger difference in the experimental group. This produced a mean between-group difference of 1.7cm

• Kinesiophobia lowered in both groups, but also more in the experimental group. A mean between-group difference of 16.3 points was noted favoring the specific exercises.

Questions and thoughts

The primary outcome significantly favored the experimental group. However, the mean between-group difference did not attain the required MCID of 20 points for the SPADI. Hence, we should refrain from emphasizing the significance of the effect. However, considering the short-term follow-up of only five weeks, a decrease in the SPADI of already 13.5 points is promising. But don’t throw the baby out with the bathwater. A large bunch of people in the experimental group did achieve a clinically important difference.

Both groups took Naproxen and the duration and the periodicity of both treatments were similar. A significant difference, however, appeared in the duration of the exercise sessions. In the control group, the duration of the exercise was 1 hour and 30 minutes, while in the intervention group, the exercises were performed in approximately 1 hour. This also means that longer is not necessarily better: the control group did way more both in terms of the number of exercises performed and in the duration of the sessions. The longer sessions did however not affect adherence in the control group.

Talk nerdy to me

A mediation analysis is a statistical approach used to study the mechanism by which an independent variable (such as an exercise program) influences a dependent variable (such as shoulder function) via an intermediary variable known as the mediator. In this study, improvements in kinesiophobia and pain intensity with movement were mediating the improvements in shoulder function. Instead of solely reducing pain (which is often thought to be the reason why people get better), the effects on shoulder function were caused by improvements in fear of movement.

This study’s success rests in its systematic approach, which included randomization, concealed allocation, and assessor blinding, which reduced biases and increased the validity of the results. Although statistically significant, some benefits fell short of the minimum clinically relevant difference (MCID), indicating that exercise programs need to be refined or intervention periods should be extended. Despite the lack of finding a true clinically important difference, an important reduction in the primary outcome at 5 weeks may be promising to study with longer follow-ups.

Take home messages

At 5 weeks after performing specific shoulder exercises the statistically significant differences favored the experimental intervention. Yet, no clinically significant between-group difference was noted for the primary outcome. Within-group improvements showed that a large part of the people in the experimental group however did attain relevant improvements beyond the MCID. Longer follow-ups may be needed to determine whether this also leads to superiority in the between-group differences.

Reference

Gutiérrez Espinoza H, Araya-Quintanilla F, Pinto-Concha S, Valenzuela-Fuenzalida J, López-Gil JF, Ramírez-Velez R. Specific versus general exercise programme in adults with subacromial impingement syndrome: a randomised controlled trial. BMJ Open Sport Exerc Med. 2023 Sep 27;9(3):e001646. doi: 10.1136/bmjsem-2023-001646. PMID: 37780129; PMCID: PMC10537831.