Introduction

“Should this exercise be so painful?” is something you’d have heard when prescribing it to patients with rotator cuff-related shoulder pain (RCRSP). Some pain with exercise is okay but mostly we are reluctant to make it very hard to bear. The opposite is true in this study where the intent was to investigate the potential benefits of painful exercise for chronic RCRSP. The benefits of resistance exercise for this condition have been studied extensively and therefore it is used in rehabilitation of RCRSP to increase load capacity and tolerance in the shoulder. Resistance exercise also induces endogenous hypoalgesia and activates the top-down pain inhibitory mechanisms. In the systematic review and meta-analysis by Smith in 2017, moderate evidence indicated that painful exercise could potentially have a benefit over pain-free exercises in the short term. It is hypothesized that the more pain increases in an exercise bout, the more pain inhibition may occur.

Methods

This study was a feasibility study with four main aims. The first was to look at adherence and adverse effects. The second was to study the time needed to collect the data. As a third objective, the study sought feedback from the participants and physiotherapists. The fourth aim was to examine the effects of painful exercise for chronic RCRSP relief.

Eligible participants were between the ages of 18 and 65 years. They had shoulder pain in the anterolateral shoulder region for at least 3 months. At rest, their maximum pain level was 2/10 on the verbal NRS. At least 3 positive tests out of the following were required:

• Pain during Neer test
• Pain during Hawkins-Kennedy test
• Pain with Jobe test
• Painful arc between 60° and 120°
• Pain or weakness during external rotation resistance test

Every physical therapy appointment lasted roughly 30 minutes, including 15-20 minutes of exercise treatment (exercising through pain) and 10-15 minutes of manual therapy (concentrating on stretching the posterior soft tissues of the shoulder).

A total of nine supervised exercise sessions were held over a 12-week period. Every week for the first five weeks, a supervised exercise session was held and two non-supervised home exercise sessions were completed. The remaining sessions were spread over the following 7 weeks, where 3 non-supervised sessions were scheduled in the unsupervised weeks.

The treating physiotherapist could choose 4 exercises to prescribe for each individual participant and these were chosen from a list of possible exercises. They included the following:

• Press up on a chair, arms in extension and external rotation, knees flexed at 90°, full bilateral foot support
• Press up on a chair, bilateral tip-toe support
• Press up on a chair, unilateral foot support (e.g. 1 leg extended)
• Press up on a chair, unilateral tip-toe support
• Press up on a table, bilateral tip-toe support
• Press up on a table, unilateral tip-toe support
• Press up on a table, no foot support
• Press up on kitchen counter, bilateral tip-toe support
• Press up on kitchen counter, unilateral tip-toe support
• Press up on kitchen counter, no foot support

Push out

• Push out on the wall with both arms
• Push out on the wall with the affected arm (not shown in the picture)
• Push out on the wall with the affected arm, with contralateral leg extended
• Push out on the kitchen counter with both arms
• Push out on the kitchen counter with the affected arm
• Push out on the kitchen counter with the affected arm, with contralateral leg extended
• Push out on the table/desk with both arms
• Push out on the table/desk with the affected arm
• Push out on the table/desk with the affected arm, with contralateral leg extended
• Push out on lower height/stairs
• Push out on the floor

External rotation against the wall

• External rotation on side plank, standing against the wall (distance 1 foot)
• Same exercise as 2, support on one foot
• Increase foot distance to the wall (2 feet)
• Same exercise as 3, support on one foot
• External rotation side plank on kitchen counter/back of a sofa
• Same exercise as 5, support on one foot
• External rotation side plank on table/desk
• Same exercise as 7, support on one foot
• External rotation side plank in horizontal position (floor, bed…)
• Same exercise as 9, support on one foot

Overhead rollover

• Glide on the wall with both arms, little* distance from the wall (not shown in the picture)
• Glide on the wall with both arms, little distance from the wall, one leg extended (not shown in the picture)
• Glide on the wall with both arms, bigger* distance from the wall
• Glide on the wall with both arms, bigger distance from the wall, one leg extended
• Glide on the wall with the affected arm, little distance from the wall
• Glide on the wall with the affected arm, little distance from the wall, one leg extended
• Glide on the wall with the affected arm, bigger distance from the wall
• Glide on the wall with the affected arm, bigger distance from the wall, one leg extended
• Glide on the kitchen counter with both arms
• Glide on the kitchen counter with the affected arm
• Glide on the table/desk with both arms
• Glide on the table/desk with the affected arm
• Glide on the floor from knee support with both arms
• Glide on the floor from knee support with the affected arm

Same progressions can be made using a foam roll on the wall (or a pillow cover at home)

*Little distance corresponded to 1 foot, bigger distance corresponded to 2 feet

Exercise with graded elastic band

• External rotation at 0° of abduction
• External rotation with graded elastic band (10 kg), elbow flexed
• External rotation with graded elastic band (15 kg), elbow flexed
• External rotation with graded elastic band (20 kg), elbow flexed
• External rotation with graded elastic band (25 kg), elbow flexed

External rotation at 90° of abduction

• External rotation in overhead position with graded elastic band (10 kg), elbow flexed
• External rotation in overhead position with graded elastic band (15 kg), elbow flexed
• External rotation in overhead position with graded elastic band (20 kg), elbow flexed
• External rotation in overhead position with graded elastic band (25 kg), elbow flexed

Horizontal adduction

• Horizontal adduction with graded elastic band (10 kg), elbow flexed
• Horizontal adduction with graded elastic band (15 kg), elbow flexed
• Horizontal adduction with graded elastic band (20 kg), elbow flexed
• Horizontal adduction with graded elastic band (25 kg), elbow flexed

During the exercises, a pain level between 4 and 7 on a verbal NRS was required. Of the four exercises, one had to be performed in the pain-evoking direction, while the other three exercises were performed in a non-painful direction but it was made sure that they were painful by adding resistance.

The pain during exercise was lowered during the last three weeks of the program. This was done to “allow the patient to exercise in a less painful range after neuromuscular adaptations occurred in the previous phase”.

Results

Twelve participants were included in the study and they were on average 50 years old. They had symptoms for about 6.5 months, and in most of the cases, the dominant arm was affected.

Considering the first objective, 88% of the participants adhered to at least 7 out of 9 supervised exercise sessions, while this number dropped to 50% of the participants who completed at least 22 out of 27 non-supervised home exercise sessions.

The physiotherapists indicated that it was difficult to provide 4 painful exercises throughout the study course. Some participants recovered very quickly, while others became demotivated from the pain becoming unbearable. Considering the compliance to the home exercises, the advice to perform the exercises with a rest day in between was often ignored or sports participation was increased, influencing the pain levels.

The patient-reported outcome measures of the adherent participants revealed that 3 out of 8 reached a significant reduction in SPADI, with the difference exceeding the MCID of 20 points. One of them had a significant change below the MCID.

Questions and thoughts

A combination of 3 out of 5 shoulder tests had to be positive. Was it to confirm the presence of RCRSP or was it to include a structured and standard shoulder examination?

How did they inform P about the why of exercising into pain for RCRSP? Did they receive an explanation that exercising through pain may actually help them improve? Because, you’d have to motivate someone really good to get them going through pain, but then this might give a positive expectation which may confound the effects. So very interesting to see further research on this topic.

The loss to follow-up was large, and this may question the feasibility of the program. Was it too intense? Should more supervision be included?

Talk nerdy to me

This was a feasibility study, thus there was no randomization, and no blinding occurred. Yet, what can it teach us and what are first results? The study rationale is supported by the systematic review of Smith et al., 2017, who found that painful exercises provide a small but considerable advantage over painless activities. But they also indicated that there is no definite advantage of one treatment over another in the medium and long term. They thus concluded that successful outcomes do not necessarily require the presence of pain during therapeutic exercise for chronic musculoskeletal pain.

The exercises were chosen pragmatically, from a list of 8 possible exercises. It was not mentioned how the exercises were chosen. Yet, avoiding a one-treatment-fits-all is better than doing so, as this replicates real practice more closely.

The adherence to the study was measured in only 8 participants while 12 were included. Therefore, the adherence will likely be much less than the reported 88 and 50%. It seems that it is hard to motivate participants to continue their exercises into pain.

Not every participant did reach the predetermined pain range of NRS between 4 and 7/10. Considering the mean of pain in four exercises for every supervised session in the first 9 weeks, four patients (57%) trained between 4 and 7 on verbal NRS, while three (43%) did not reach this range. This is of concern for the validity of the study procedures because the intent of what was studied (painful exercise for chronic RCRSP) was not achieved. As such, it could be that there are subgroups of participants who are able to bear the pain, while others may be reluctant to proceed.

The effect of exercising into pain is studied, but it is supplemented with manual therapy. This is more than a supplement as it took half of the treatment time. It would be better to refer to the “effect of exercising into pain combined with manual therapy”. The manual therapy was focused on stretching of the posterior soft tissues of the shoulder, but no further details were given.

Take home messages

This was a first step to determine the feasibility of exercising into pain for chronic RCRSP. The study found that it was hard to conduct and patients either recovered quickly or resided from exercising because of pain becoming unbearable. It may be possible that the manual therapy provided confounded the results. The adherence was not good for the unsupervised sessions and the adherence of the supervised sessions was only investigated in 8 of the 12 included participants. The trial had only one group and the procedures were not randomized, it seems that this is thus the only way to conclude the possible effectiveness of exercising into pain for chronic RCRSP.

Reference

Cavaggion C, Juul-Kristensen B, Luque-Suarez A, Voogt L, Wollants G, Ó Conaire E, Struyf F. Exercise into pain in chronic rotator cuff related shoulder pain: a prospective single-group feasibility study. BMJ Open. 2023 Oct 6;13(10):e070698. doi: 10.1136/bmjopen-2022-070698. PMID: 37802620; PMCID: PMC10565173.