Introduction

Joint hypermobility is advantageous in throwing sports where high flexibility is required. Yet patients with hypermobility spectrum disorder are at greater risk of developing shoulder symptoms at some point in their life. Some earlier studies found exercise beneficial in increasing strength and tendon stiffness around the shoulder joint, however specific exercise recommendations are lacking. In an earlier trial, high-load shoulder strengthening was found feasible and safe for people with HSD and long-lasting shoulder symptoms. But to discover whether high-load strengthening in hypermobile shoulders was able to improve self-reported shoulder function, this RCT was set up!

Methods

In this trial, a low-load strengthening program was compared against a high-load strengthening program for self-reported functioning in patients with hypermobility spectrum disorder (HSD). This was defined as

• Generalized HSD (G-HSD) is defined using a Beighton score cut off ≥ 5/9 for females up to the age of 50 years, and ≥ 4/9 for those > 50 years and all males, OR
• Historical HSD (H-HSD) if the Beighton score is 1 point below the age and sex-specific cut-off and the 5PQ is positive (≥ 2/5 positive answers)

Next to this, participants had to have at least one of the following symptoms:

• Musculoskeletal pain in at least one shoulder for at least 3 months.
• Recurrent joint dislocations or joint instability without a reported history of trauma is defined as (a) a minimum of three atraumatic dislocations in the same shoulder, (b) a minimum of two atraumatic dislocations in two different joints (a minimum of one in the shoulder) occurring at different times, and/or (c) medical confirmation of joint instability in at least two joints (a minimum of one in the shoulder).

Participants in the high-load group received 5 exercises to perform 2 times a week under supervision and once a week at home. The exercises were performed using dumbbells up to 15 kg, with individually adjusted loads. The exercises in the first 3 weeks were performed at an increasing load from 50% to 70%, to 90% of 10RM. Thereafter, from weeks 4-9 loads were increased to 10RM, and in weeks 10-15 sets were performed using loads of 8RM. In the picture here below, you can see the detailed description of the high-load strengthening program. The exercises performed were the following: Side-lying external rotation in neutral, prone horizontal abduction, prone external rotation at 90° shoulder abduction, supine scapular protraction, and seated scaption.

In the low-load strengthening group, the intention was to mimic standard care practice in Denmark. Here self-training is regularly prescribed 3 times a week. In the study, exercises were introduced to the patients and they were supervised at weeks 5 and 11 when the new exercises were conducted. The program included nine shoulder exercises:

• phase 1 (isometric), posture correction;
• phase 2 (isometric), shoulder abduction, shoulder internal and external rotation with 90° flexion at the elbow joint against a wall, and standing weight-bearing in the shoulders against a table;
• phase 3 (dynamic with a yellow Theraband), shoulder abduction, shoulder internal and external rotation at 90° flexion at the elbow joint, and four-point kneeling with single-arm raising.

The primary outcome was the Western Ontario Shoulder Instability Index (WOSI), which is a self-reported measure of shoulder function. This was measured in the short term after sixteen weeks. The questionnaire consists of 21 questions rated on a 0-100 scale where 0 represents the absence of shoulder limitations. Adherence to the exercise program was measured. When less than 32 of the 48 exercise sessions were completed, the participant was classified as being non-adherent.

Results

A hundred participants were randomly assigned to one of both strengthening programs and data of 93 was available at 16 weeks. Seventy-nine percent of participants were women and the mean age of the sample was 37.8 years. Their mean Beighton score was 5.8. In total, 48 exercise sessions were to be followed and 67 patients were classified as adherers (34 and 33 in the high- and low-load groups respectively).

The primary analysis revealed that the high-load strengthening in hypermobile shoulders led to greater improvements in shoulder function compared to the low-load program. The mean difference was -174.5 points (adjusted for baseline WOSI score, age, sex, and clustering around the physiotherapy clinic), which lies under the minimal important difference. When the per-protocol analysis was done, the mean difference between both groups attained -250.7 points. These findings were supported by sensitivity analyses, see questions, and thoughts.

When clinically relevant outcomes were evaluated, it was reported that nearly ⅔ of patients in the high-load program attained a relevant outcome, compared to approximately half of patients in the low-load group. A change of at least 252 points in WOSI was defined as a clinically relevant outcome. The adjusted number needed to treat was 3, with a narrow confidence interval ranging between 2 and 7 patients.

Questions and thoughts

By using the WOSI questionnaire, this study used a responsive, valid questionnaire that is sensitive to change and has a high test-retest reliability. The minimal important difference was reported to range between 10.4% and 14%, thus between 218.4 and 294 points. The per-protocol analysis revealed a mean improvement of 250.7 points, which lies in between these previously reported values. Unfortunately, the intention to treat analysis did not attain the minimal important difference. We discuss this hereunder in “Talk Nerdy to Me”.

Are you familiar with the thought of using low-load shoulder exercises in those with hypermobile/unstable shoulders at first? Are you scared that these patients would more easily dislocate their shoulders during heavy exercise? This isn’t surprising as there remains uncertainty about the safety and effectiveness of high-load shoulder exercise in hypermobility disorders. In fact, some guidelines recommend against high-load strengthening!! In 2020, Liaghat et al. challenged this recommendation and found high-load strengthening in hypermobile shoulders to be safe and feasible in their feasibility trial. This study adds further evidence to ignore the recommendation against high-load strengthening in hypermobile shoulders. Luckily, the reported adverse events were all minor. More adverse events occurred in the high-load strengthening group and these were muscle soreness and headache. So no real issues. In both groups, a minority had a new subluxation or dislocation: 3 subluxations occurred in the low-load group, one in the high-load group, and 1 dislocation occurred in the high-load group. Crazy enough, these adverse events were grouped under minor adverse events. So-called major adverse events covered death, life-threatening events, disability, and permanent damage. Yet, to a patient with hypermobile shoulders, a redislocation would seem a major concern in my opinion.

Personally, I wouldn’t call shoulder hypermobility the bogeyman. However, to avoid shoulder symptoms arising from high shoulder loads when going through a hypermobile range, I think it is crucial to maximize control over that range. High-load strengthening exercises like the ones studied here may be valuable to improve the function of the stabilizing muscles surrounding the shoulder joint.

Talk nerdy to me

The participant was classified as non-adherent when less than 32 out of 48 sessions were completed. In fact, thirty-three participants didn’t complete at least 32 sessions. Was the program too heavy that one-third of participants didn’t succeed to complete more than two-thirds of the exercise program? Or can we turn it the other way, would it be sufficient to complete less than the predefined number of 48 sessions? The sensitivity analysis accounted for both. In the intention-to-treat analysis, all randomized participants were followed. This analysis revealed that the mean between-group improvement in WOSI yielded -198.7 points. In the per-protocol analysis, where only those who followed the complete trial (the adherers) were analyzed, the mean between-group difference in WOSI was -250.7. It is very logical that in the latter, a higher between-group difference was found. A common reason for dropping out is that people aren’t improving. The adherent participants are likely those who see improvements in their symptoms and are or become motivated to complete the program. When we look at the more critical analysis – the intention-to-treat analysis where baseline values are carried forward based on the assumption that dropped-out participants will likely return to their baseline situation – we see that the change between the groups did not attain the minimal important difference. So is the high-load strengthening program really superior to the low-load one? At least not at the short-term follow-up of 16 weeks. However, we can derive some lessons for clinical practice, where we work in not-so-strictly controlled environments like in RCTs. When your patient is adherent to the exercise sessions, you may expect greater improvements with the high-load strengthening program, as derived from the per-protocol analysis.

Power calculation was based on the findings from the feasibility study that was conducted in 2020. This way, the authors could make use of representable data to ensure a good sample size calculation was done. Furthermore, the paper was published with open access, and with a very detailed protocol and statistical analysis plan. When looking at these files, they planned to publish graphs that are lacking in the final paper.

The authors deserve a huge thumbs up for the use of the Consensus on Exercise Reporting Template (CERT), which is lacking in many clinical exercise trials. This is a way to ensure transparency and improve trial interpretation so that effective exercise interventions can more easily be implemented into clinical practice.

Take home messages

It seems that high-load strengthening in hypermobile shoulders led to a greater improvement in shoulder function. However, the mean difference did not attain the minimal important difference. These results were obtained in a sample that consisted of mostly women (79%), so generalizability to males is limited. Those who followed the predefined protocol achieved greater improvements with the high-load strengthening protocol. Little adverse events (headache and muscular soreness) occurred in this study. Only 1 patient had a subluxation event in the high-load strengthening group compared to 3 in the low-load group. One patient in the high-load strengthening group dislocated their shoulder compared to none in the low-load group.

Useful content

Reference

Liaghat B, Skou ST, Søndergaard J, Boyle E, Søgaard K, Juul-Kristensen B. Short-term effectiveness of high-load compared with low-load strengthening exercise on self-reported function in patients with hypermobile shoulders: a randomised controlled trial. Br J Sports Med. 2022 Jun 1:bjsports-2021-105223. doi: 10.1136/bjsports-2021-105223. Epub ahead of print. PMID: 35649707.