Shoulder Pain Relief: What Works and What the Evidence Supports

Most shoulder pain responds to exercise-based rehabilitation, and the diagnosis matters less than many people assume.

The shoulder is the most mobile joint in the body. The glenohumeral joint allows movement through approximately 180 degrees in multiple planes, achieving this range by sacrificing the bony stability that the hip or knee rely on. A shallow socket (the glenoid), a loose capsule, and a group of four rotator cuff muscles provide the dynamic stability that keeps the humeral head centered during this motion. The scapula, sitting on the rib cage and controlled by its own set of muscles, must rotate, tilt and protract in coordination with the arm. When any part of this system is not functioning well, the shoulder compensates, and pain follows.

This mobility-stability tradeoff is the key to understanding most shoulder pain. The hip has a deep socket and strong ligaments; it sacrifices range for inherent stability. The shoulder goes the other direction: maximal range, minimal bony constraint, and near-total dependence on muscular control to keep the joint centered under load. When those muscles are weak, fatigued or poorly coordinated, the joint does not dislocate, but it does develop compensatory movement patterns that produce pain over time.

Shoulder pain is the third most common musculoskeletal complaint in primary care. The most frequent diagnoses include rotator cuff tendinopathy, subacromial impingement, adhesive capsulitis (frozen shoulder), and acromioclavicular joint pathology. What is increasingly recognized is that these labels overlap considerably, and the specific diagnosis is less important for treatment decisions than many patients expect.

What Is Happening

The concept of “rotator cuff related shoulder pain” has emerged as an umbrella term that captures the reality of most non-traumatic shoulder pain: the rotator cuff muscles and tendons are not functioning optimally, the scapular stabilizers may be contributing, and the result is pain with arm elevation and use [3]. The specific anatomical structure that is the primary pain generator is often difficult to identify, and, importantly, the treatment is similar regardless.

This is not nihilism. It is an evidence-based recognition that exercise-based rehabilitation targeting the rotator cuff and scapular stabilizers works across diagnostic categories. A systematic review of 36 RCTs confirmed that exercise is effective for pain and function in shoulder impingement, with effects maintained at follow-up [1]. A separate review found consistent evidence for exercise in rotator cuff tendinopathy specifically [4].

The reason exercise works across these diagnoses is that the underlying problem is frequently the same: the muscles that center the humeral head and coordinate scapular movement are not doing their job adequately. Whether the result is labeled tendinopathy, impingement, or non-specific shoulder pain, the treatment principle holds: restore the strength and coordination of the rotator cuff and scapular stabilizers, and the pain resolves in most cases.

There are exceptions. Adhesive capsulitis (frozen shoulder) involves progressive capsular contracture and has a distinct natural history: a painful phase, a stiffening phase, and a thawing phase over 1 to 3 years. It benefits from gentle mobilization but does not respond to standard strengthening in the same way. Acromioclavicular joint pain, felt at the top of the shoulder, has a different mechanism and may require different management. Recognizing the exceptions is important; overgeneralizing is not the same as recognizing patterns.

The Inflammation Question

Many patients with shoulder pain are told they have “inflammation” in the shoulder, and this shapes their expectations. They expect anti-inflammatory treatments to work: ice, NSAIDs, cortisone injections. The relationship between shoulder pain and inflammation is more nuanced than this framing suggests [6].

Acute tendon injuries involve an inflammatory response, which is a necessary part of healing. Chronic tendinopathy, which is what most patients with shoulder pain lasting more than a few weeks have, involves a different tissue state: disordered collagen, increased vascularity, and neurochemical changes, but not necessarily active inflammation in the traditional sense. This matters because treatments that target inflammation (particularly corticosteroid injection) may not be addressing the actual tissue pathology in chronic presentations.

Corticosteroid Injection: Short-Term Help, No Long-Term Advantage

Corticosteroid injection into the subacromial space is one of the most commonly performed procedures for shoulder pain. It works in the short term: most patients experience reduced pain for 6 to 12 weeks.

However, a Lancet systematic review of 41 RCTs found that corticosteroid injections for tendinopathy provided short-term pain relief but were associated with worse outcomes in the intermediate and long term compared to other treatments or no injection [2]. The injections suppress inflammation temporarily but do not address the underlying mechanical problem, and there is concern that repeated injections may weaken tendon tissue over time.

The reasonable use of injection is as a bridge: to reduce pain enough to allow participation in rehabilitation when pain is so severe that exercise-based treatment cannot begin. It is not a standalone treatment, and its short-term benefit should not be mistaken for resolution.

Surgery: When and When Not

The CSAW trial, published in the Lancet in 2018, randomized 313 patients with subacromial shoulder pain to arthroscopic subacromial decompression, sham surgery, or no treatment. At one year, real surgery was no better than sham surgery [5]. This is one of the clearest results in orthopedic surgery: the most commonly performed shoulder operation for impingement does not outperform a placebo procedure.

Surgery remains appropriate for specific conditions: large rotator cuff tears with significant weakness (particularly in younger or active patients where tendon retraction may progress), failed conservative management after 6 months of consistent effort, or structural problems (unstable labral tears, acromioclavicular joint pathology with a clear mechanical cause) that have a specific surgical indication. But for the common presentation of shoulder pain with arm elevation, exercise is the first and often the only treatment needed.

Exercises to Try

External rotation with a band. Stand with the elbow bent to 90 degrees and tucked to the side. Hold a resistance band in both hands. Rotate the affected arm outward against the band’s resistance. Keep the elbow stationary; if the elbow drifts away from the body, the band is too heavy. Three sets of 15. This is the foundational rotator cuff exercise, targeting infraspinatus and teres minor. It restores the external rotation force that centers the humeral head during arm elevation.

Side-lying external rotation. Lie on the uninvolved side. Hold a light weight (1 to 2 kg), elbow at 90 degrees. Rotate the forearm toward the ceiling. Pause at the top for 2 seconds. Lower slowly over 3 seconds. Three sets of 12. This adds gravity as resistance and targets infraspinatus and teres minor through a greater range than the banded version. Keep a rolled towel between the elbow and the body to maintain consistent positioning.

Prone Y, T and W raises. Lie face down on a bench. For Y: lift arms at 45 degrees from the body, thumbs up. For T: lift arms straight to the side, thumbs up. For W: elbows bent, squeeze shoulder blades together and rotate the forearms up. Each targets different scapular stabilizers (lower trap for Y, middle trap for T, external rotators with retraction for W). Three sets of 10 each. Start without weight; add 0.5 to 1 kg when all sets are pain-free.

Loaded flexion. Stand holding a light dumbbell. Raise the arm forward in the scapular plane (about 30 degrees forward of directly to the side), thumb pointing up, to shoulder height or slightly above. Lower slowly over 3 seconds. Three sets of 12. Progress weight gradually. This loads the supraspinatus and deltoid in a functional range while the thumb-up position externally rotates the humerus, reducing subacromial compression.

What Imaging Can and Cannot Tell You

Many patients with shoulder pain expect an MRI to provide a definitive answer. In practice, imaging can complicate the picture as much as clarify it. Rotator cuff tendon changes, including partial tears, are common in people over 40 who have no shoulder pain at all. Finding a partial tear on MRI does not necessarily mean the tear is causing the pain, and it does not necessarily mean surgery is needed.

Imaging is appropriate when symptoms persist despite 8 to 12 weeks of consistent rehabilitation, when there is sudden significant weakness after an injury (suggesting a substantial acute tear), or when surgery is being actively considered and the surgeon needs to assess the extent of damage. For the initial presentation of shoulder pain without red flags, clinical examination is usually sufficient to guide the first phase of treatment. A clinician who can assess range of motion, strength in specific positions, and provocative tests can identify the likely problem category and begin the correct rehabilitation without an MRI.

The risk of early imaging in non-urgent shoulder pain is not physical; it is psychological and decisional. An MRI report describing “tendinosis,” “partial tear,” or “bursitis” can lead a patient to believe their shoulder is more damaged than it is, reduce their confidence in exercise-based treatment, and push both patient and clinician toward interventions (injection, surgery) that the evidence does not support as first-line treatments for these findings.

When to See a Doctor

See a doctor or physical therapist if:

  • There is significant weakness, particularly sudden onset after an injury (possible rotator cuff tear)
  • Range of motion is rapidly decreasing (possible adhesive capsulitis)
  • Night pain is severe and persistent
  • Pain has not improved after 8 to 12 weeks of consistent exercise
  • There is numbness or tingling in the arm (may indicate cervical radiculopathy rather than a shoulder problem)

When to see a doctor. If joint pain follows a significant injury, is accompanied by swelling, giving way, locking, or loss of range of motion, or if you have neurological symptoms (numbness, tingling, weakness), see a doctor or physical therapist. They can assess whether imaging or specialist input is appropriate.

Frequently asked questions

What is the best treatment for shoulder pain?

Exercise-based rehabilitation targeting the rotator cuff and scapular stabilizers has the strongest evidence for most common causes of shoulder pain. A 2017 systematic review of 36 RCTs found that exercise is effective for pain and function across rotator cuff tendinopathy, impingement, and non-specific shoulder pain.

Should I get a cortisone injection for shoulder pain?

Corticosteroid injection can reduce pain in the short term (6 to 12 weeks) and may help you participate more fully in rehabilitation. However, multiple studies show that injection does not improve long-term outcomes beyond exercise alone, and repeated injections may weaken tendon tissue.

Do I need an MRI for shoulder pain?

Not for most initial presentations. Clinical examination by a skilled clinician can identify the likely problem. MRI is appropriate when symptoms persist despite rehabilitation, when there is significant weakness suggesting a substantial tear, or when surgery is being considered.

How long does shoulder pain take to resolve?

Most shoulder pain from rotator cuff tendinopathy or impingement improves meaningfully within 6 to 12 weeks of consistent exercise. Frozen shoulder has a longer timeline of 1 to 3 years. Complete rotator cuff tears that require surgery have a 4 to 6 month rehabilitation period.

Sources and further reading

  1. Steuri R et al. — Effectiveness of conservative interventions including exercise, manual therapy and medical management in adults with shoulder impingement: a systematic review and meta-analysis

    Br J Sports Med, 2017. 36 RCTs, 2,355 participants. DOI: 10.1136/bjsports-2016-096515

  2. Coombes BK et al. — Efficacy and safety of corticosteroid injections and other injections for management of tendinopathy: a systematic review

    Lancet, 2010. 41 RCTs. DOI: 10.1016/S0140-6736(10)61160-9

  3. Lewis J — Rotator cuff related shoulder pain: assessment, management and uncertainties

    Man Ther, 2016. Review. DOI: 10.1016/j.math.2016.03.009

  4. Littlewood C et al. — Exercise for rotator cuff tendinopathy: a systematic review

    Physiotherapy, 2012. 9 trials. DOI: 10.1016/j.physio.2011.08.002

  5. Beard DJ et al. — Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW)

    Lancet, 2018. 313 participants. DOI: 10.1016/S0140-6736(17)32457-1

  6. Scott A et al. — What do we mean by the term 'inflammation'? A contemporary basic science update for sports medicine

    Br J Sports Med, 2004. Review. DOI: 10.1136/bjsm.2004.011312

Sources checked January 3, 2026. This is an editorial summary, not a systematic review or individual advice.

Get new articles by email

New articles from Joint by Joint, sent to your inbox. Unsubscribe at any time.

Related articles