Overuse shoulder injuries such as rotator cuff (RC) tendinopathy are common with a prevalence estimated to be 14% in the general population of which 23% of the working population with shoulder problems are sick listed.

RC tendinopathy is a tendon-related pain in the proximal lateral aspect of the upper arm with weakness, especially during active elevation and external rotation, and painful active range of motion. The prevalence of RC tendinopathy is highest in the supraspinatus and infraspinatus tendon.

Exercise therapy is regarded as an effective intervention for symptomatic RC tendinopathy for reducing pain and disability and improving function. However, the prescription is diverse and the effectiveness of specific characteristics of exercise programs is unknown. Many contextual factors and prescription parameters, such as external resistance, training intensity and frequency, home versus supervised exercises, duration of the program, etc. have been described. Some of these prescription parameters have been extensively studied, with some conflicting results. Although some level of resistance seems to matter, as well as number of sets and repetitions, the 'optimal' level and volume are unclear. Inducing or allowing pain based on tendon loading during exercises is todays consensus in the treatment of patellar and achilles tendinopathy using a pain-monitoring model. Although a number of shoulder studies report that pain either should be avoided or allowed, not one study ever examined the influence of pain allowance versus pain avoidance during a shoulder exercise program on patient outcome in terms of pain, physical function and disability.

The purpose of this project is to examine the effect of allowing pain versus avoiding pain based on tendon loading during an exercise regimen for patients with symptomatic rotator cuff (RC) tendinopathy. This will be accomplished in a Randomized Controlled Clinical Trial, comparing the effectiveness on patient reported and objective outcomes of a "pain allowing" and "pain avoiding" exercise program, performed for 26 weeks. Our hypothesis is that allowing pain based on tendon loading during exercises would result in a better outcome in pain and function measured on SPADI (the primary outcome) compared to avoiding pain in patients with RC tendinopathy.

Aim The purpose of this project is to examine the effect of allowing pain versus avoiding pain based on tendon loading during an exercise regimen for patients with symptomatic rotator cuff (RC) tendinopathy. This will be accomplished in a Randomized Controlled Clinical Trial, comparing the effectiveness on patient reported and objective outcomes of a "pain allowing" and "pain avoiding" exercise program, performed for 26 weeks.

Background Shoulder disorders are the third most common musculoskeletal disorder with a life-time prevalence in the general population of 30%. Shoulder disorders are often persistent and recurrent, with 54% of the patients reporting on-going symptoms after 3 years. RC tendinopathy is regarded as a common source of shoulder pain with prevalence estimated to be as high as 14% in the general working-age population. About 23% of the working population with shoulder problems are sick-listed, with a potential individual productivity loss.

RC tendinopathy is a tendon-related pain in the proximal lateral aspect of the upper arm with weakness, especially during active elevation and external rotation, and painful active range of motion. The pathoanatomic understanding is extended by classifying tissue irritability and specific impairments. Tissue irritability is meant to guide intensity of treatment, and identifying specific impairments guides specific tactics used for intervention. The prevalence of RC tendinopathy is highest in the supraspinatus and infraspinatus tendon.

Exercise therapy is widely regarded as an effective intervention for symptomatic RC tendinopathy for reducing pain and disability and improving function. It is well known that tendon collagen regains formation and tensile strength faster than unstressed collagen, and that it can take 12 month or longer before it reaches full maturity and strength. However, the prescription is diverse and the effectiveness of specific characteristics of exercise programs is unknown. Many contextual factors and prescription parameters, such as external resistance, training intensity and frequency, home versus supervised exercises, duration of the program, etc. have been described and are summarized in a systematic review. Some of these prescription parameters have been extensively studied, with some conflicting results. Although some level of resistance seems to matter, as well as a number of sets and repetitions, the 'optimal' level and volume are unclear. The existing studies offer some preliminary guidance in relation to the development and application of loading in exercise programs for RC tendinopathy, however there is a gap in literature with respect to allowing or avoiding pain during shoulder exercises. Inducing or allowing pain based on tendon loading during exercises is todays consensus in the treatment of patellar and achilles tendinopathy using a pain-monitoring model.

Although a number of shoulder studies report that pain either should be avoided or allowed, not one study ever examined the influence of pain allowance versus pain avoidance during a shoulder exercise program on patient outcome in terms of pain, physical function and disability.

Hypothesis Allowing pain based on tendon loading during exercises would result in a better outcome in pain and function measured on SPADI (the primary outcome) compared to avoiding pain in patients with RC tendinopathy.

Inclusion Criteria:

• Adults aged 18 to 55 years
• Shoulder symptoms lasting for a minimum of 3 months
• Clinical diagnosis of rotator cuff (supraspinatus and/or infraspinatus) tendinopathy
• Clinical diagnosis verified by/ combined with ultrasound

Exclusion Criteria:

Patients are excluded if they have

• resting pain above 4/10 (NPRS)
• <90 degrees active elevation of the arm
• had a corticosteroid injection within the previous 12 weeks
• isolated subscapularis tendinopathy
• total rotator cuff tear
• diagnosed AC-joint pathology
• diagnosed labrum pathology
• diagnosed glenohumeral joint instability
• had prior shoulder surgery (all shoulder joints)
• diagnosed glenohumeral osteo arthrosis (OA) evaluated on x-ray, rheumatoid arthritis or periarthrosis General exclusion criteria are inability to speak or read Danish, inability to perform and maintain the physical training, or other conditions negatively influencing compliance.

Double (investigator, outcome assessor)