PRP vs. Corticosteroid Injections for Frozen Shoulder: How Do Pain and Mobility Outcomes Compare?
Published November 16, 2026
Protocol: 7 steps in this guide
Two 2026 studies point in opposite directions. A meta-analysis of seven randomized trials with 692 participants found PRP and corticosteroid injections similar for the first month, then PRP ahead on pain, range of motion and function at 3 and 6 months. A 90-person randomized trial published a few weeks later, in which every patient also had physical therapy, found the corticosteroid group did better on several measures through six months, “particularly during the freezing stage.” All groups in that trial improved. Neither study settles the question, and the American Academy of Orthopaedic Surgeons notes that frozen shoulder “tends to get better” over time.
Frozen shoulder and its stages
In frozen shoulder, also called adhesive capsulitis, the American Academy of Orthopaedic Surgeons explains on its frozen shoulder page that “the shoulder capsule thickens and becomes stiff and tight.” It describes three stages:
| Stage | What happens, per AAOS | Typical duration, per AAOS |
|---|---|---|
| Freezing | Pain slowly increases and the shoulder loses range of motion | 6 weeks to 9 months |
| Frozen | Pain may improve, but stiffness remains | 4 to 6 months |
| Thawing | Motion slowly improves | 6 months to 2 years to return to normal or near normal |
AAOS notes that frozen shoulder “generally gets better over time without surgery, although it may take up to 3 years.” It adds: “Physical therapy, with a focus on shoulder flexibility, is the primary treatment recommendation.” It also notes that frozen shoulder occurs much more often in people with diabetes. Because the condition can improve on its own, a comparison group is essential in any trial. Frozen shoulder is a different problem from shoulder arthritis, covered in whether PRP can improve symptoms of shoulder osteoarthritis, and from rotator cuff tendinopathy.
The two injections, with or without rehabilitation
A corticosteroid injection reduces inflammation. PRP is made from the patient’s own blood; how it is prepared is covered in how platelet-rich plasma is made from a blood sample. In the 2026 trial below, the injections were given together with physical therapy, so it tested injection plus therapy, not injection alone.
Pain and range of motion are separate outcomes
In these studies, pain was measured on a visual analog scale and range of motion in degrees, for movements such as raising the arm out to the side or rotating it. Function questionnaires such as SPADI and DASH combine both. A treatment can help pain more than motion, or the reverse.
The seven-trial meta-analysis
Peng and colleagues, Journal of Back and Musculoskeletal Rehabilitation (2026), Platelet-Rich plasma vs corticosteroid for primary frozen shoulder, pooled 7 randomized trials with 692 participants, judged at overall low risk of bias. Its findings:
| Time point | Result |
|---|---|
| Up to 1 month | Comparable improvement in pain and function |
| 3 months | PRP better: pain (mean difference in VAS change -0.62), abduction (+16.68°), external rotation (+12.22°), internal rotation (+8.35°), SPADI (-11.52) |
| 6 months | PRP better: pain (VAS change -1.06), SPADI (-17.23), DASH (-8.46) |
| Complications | Rare and similar: 0.86% with PRP vs 0.87% with steroid |
The authors judged the quality of evidence low to moderate, said improvements at 6 months “are highly likely to be clinically significant” with lower likelihood at 3 months. They added: “These findings should be interpreted cautiously.”
The 90-person three-arm trial
Hsieh and colleagues, Archives of Physical Medicine and Rehabilitation (2026), Platelet-Rich Plasma Versus Corticosteroid Injection in Patients With Primary Frozen Shoulder, randomized 90 people with primary frozen shoulder to PRP plus physical therapy, corticosteroid plus physical therapy, or saline plus physical therapy, 30 in each group. Everyone received two ultrasound-guided injections into both the shoulder joint and the bursa above it. Results:
- The corticosteroid group improved significantly more on SPADI, a second disability questionnaire, pain during activity, several active and passive range-of-motion measures, and self-assessment.
- “The therapeutic effect of CS was evident at 1 month and persisted through the 6-month follow-up.”
- Nonetheless, “all 3 treatment groups exhibited significant postintervention improvement.”
The authors concluded that “CS injections yielded superior outcomes compared with PRP, particularly during the freezing stage of frozen shoulder.”
Why the results differ
The two studies differ in ways that can explain discordant results. The trial added physical therapy to every group, used a specific two-site injection technique, and included a saline-plus-therapy group, which shows how much patients improved with physical therapy plus an inactive injection, follow-up and time, not the effect of physical therapy alone. The meta-analysis pooled trials with different PRP preparations, steroid doses, injection sites and stages of frozen shoulder. The trial was published after the meta-analysis, so it was not part of it. Neither study is wrong; they answer slightly different questions. How to read comparisons like these is covered in what clinically meaningful improvement means in a joint pain study.
Deciding without a clear winner
A reasonable conversation with a clinician covers which stage the shoulder is in, whether physical therapy is part of the plan, whether diabetes affects the choice of a steroid, cost, and how quickly relief is wanted. PRP costs and coverage are covered in what PRP treatment costs and what drives the price, and PRP risks in the main risks of a PRP injection. The same injection comparison for the wrist is covered in PRP vs. corticosteroid injections for carpal tunnel syndrome.
For frozen shoulder, a 2026 meta-analysis favored PRP at 3 and 6 months while a 2026 trial combining injections with physical therapy favored a corticosteroid, especially in the freezing stage, so neither injection can be called the clear winner and the stage of the condition and the rehabilitation plan matter as much as the injection chosen.
Cells linked to this one
This guide leads on to
- C8 Can PRP Improve Symptoms of Shoulder Osteoarthritis?
- B1 How Is Platelet-Rich Plasma Made From a Blood Sample?
- I9 What Does Clinically Meaningful Improvement Mean in a Joint Pain Study?
- B11 What Does PRP Treatment Cost, and What Drives the Price?
- B10 What Are the Main Risks of a PRP Injection?
- H4 PRP vs. Corticosteroid Injections for Carpal Tunnel Syndrome: What Do Trials Show?
Add a remark