Ross Hauser, MD; Danielle Matias, PA-C; Demi Foster, APRN
Shoulder arthritis can make reaching overhead, putting a hand behind the back, sleeping, and playing golf increasingly difficult. People considering shoulder replacement often ask whether injections could reduce their pain or help them remain active without an operation.
At Caring Medical, our clinical perspective considers the supporting ligaments, tendons, and joint capsule alongside the arthritic joint surfaces. Prolotherapy is an injection treatment intended to stimulate a local repair response in painful supporting tissues. Our published review discusses its use across several musculoskeletal conditions. That rationale does not establish that injections rebuild an advanced arthritic shoulder or outperform replacement. (1)
This article preserves the original shoulder research and imaging examples while adding a case presented by Danielle Matias, PA-C. The case illustrates an important distinction: a person can report less pain while substantial joint damage, limited movement, and candidacy for shoulder replacement remain.
Shoulder osteoarthritis and rotator cuff disease are different problems
The glenohumeral joint is the ball-and-socket joint between the upper arm bone and shoulder blade. Osteoarthritis involves changes in its cartilage, underlying bone, and other joint tissues. X-rays may show narrowing of the joint space, bone spurs, and altered joint shape.
The rotator cuff is a group of muscles and tendons that helps move and stabilize the shoulder. Tendinopathy means a tendon disorder, often involving degenerative changes. The acromioclavicular, or AC, joint is a separate joint at the top of the shoulder.
These conditions may coexist, but evidence for a rotator cuff injection cannot automatically be applied to bone-on-bone glenohumeral arthritis. This distinction matters because much of the research discussed in the original article studied tendons or other shoulder pain rather than advanced ball-and-socket arthritis.
Pain severity also cannot be read directly from an X-ray. A study of 515 patients found that radiographic measures of glenohumeral arthritis did not consistently correspond to patient-reported pain and function. That does not make the imaging unimportant; it means the structural findings and the person’s limitations answer different questions. (2)
What are we seeing in this image?
A shoulder x-ray showed a large bone spur. This is an X-ray of a patient who was only in their ’30s. The patient suffered from significant shoulder pain and a limited range of motion. In this image, the shoulder cartilage is fairly well preserved, but a large osteophyte or bone spur is present. This is a situation where Prolotherapy can be helpful with the elimination of the patient’s pain; the bone spur causing restricted movement may require surgery.
Pain arising from surrounding tissues and a bony obstruction are different treatment targets. A reduction in pain would not establish that the spur had disappeared. Depending on its location, size, and contribution to symptoms, a spur may remain relevant to a surgical discussion even if an injection helps pain.
Our shoulder bone-spurs article examines that distinction in more detail.
Our published research: 94 patients with chronic shoulder pain
Our 2009 retrospective study included 94 patients with an average of 53 months of unresolved shoulder pain. Participants were questioned an average of 20 months after their last treatment about pain, stiffness, crunching sensations, daily activities, and quality of life. The historical study used quarterly treatments; that is a description of the research, not a schedule for another patient. (3)
Patients reported improvements in pain, stiffness, and crepitation—the grinding or crunching sensation with movement. Ninety-seven percent reported some pain relief, and the paper reported improvements in sleep, exercise ability, emotional symptoms, and disability in more than 82%. The sample included people previously told there were no other treatment options or that surgery was their only option. (3)
These observations help explain our interest in Prolotherapy. However, the study was uncontrolled and questionnaire-based, with potential recall and selection bias. It was not a trial confined to advanced glenohumeral arthritis. Its percentages should not be presented as the likelihood that a patient with bone-on-bone arthritis will improve, regenerate cartilage, or avoid replacement.
Rotator cuff tendinopathy: dextrose versus control injections
A randomized trial compared dextrose injections at painful tendon attachments with two saline-based control approaches in people with rotator cuff tendinopathy receiving physical therapy. At nine months, 59% in the dextrose group maintained the specified pain improvement, compared with 37% after saline at the attachments and 27% after superficial saline. The difference was statistically significant against the superficial control but not against the attachment-site saline group. (4)
This is more precise than describing the trial as showing superiority over every control. It also concerns rotator cuff pain, not restoration of an arthritic joint surface.
Refractory rotator cuff disease: one-year observational results
A retrospective case-control study analyzed 57 treated patients and 53 controls over one year. The authors reported improvements in pain, disability, strength, and active shoulder movement with Prolotherapy. Because treatment was not randomized, differences between the groups and other influences could have contributed. (5)
A published chronic shoulder pain case
A separate report described a middle-aged man receiving injections at several shoulder locations, including the glenohumeral and AC joints, subacromial space, and supraspinatus insertion. After three sessions, he reported approximately 90% less pain and a return to normal activity. A single case can describe an outcome, but it cannot establish a success rate or prove that the treatment would avoid surgery for other patients. (6)
Acromial attachment pain and AC-joint disease
A single-arm prospective study used ultrasound-guided 15% dextrose injections for acromial enthesopathy and AC-joint arthropathy. Twenty of 31 patients reported substantial pain reduction, without reported adverse effects. There was no comparison group. These findings concern structures around the top of the shoulder, not proof of benefit for advanced glenohumeral osteoarthritis. (7)
Orthobiologic review: an important evidence gap
A 2020 preprint study was published in the Journal of back and musculoskeletal rehabilitation in 2021 included 20 studies of Prolotherapy, PRP, or cell-based approaches. Results were mixed, and no glenohumeral osteoarthritis studies met its inclusion criteria. Thus, its shoulder findings could not establish effectiveness for glenohumeral arthritis. (8)
The original discussion noted that some studies using repeated injections had more sustained results than single-injection studies. This is worth investigating, but comparing different studies cannot isolate the effect of injection number when diagnoses, solutions, targets, and other treatments also differ. It does not establish that a single injection never helps or that more injections are universally superior.
Updated 2025–2026 research: results remain mixed
A 2025 review reported improvements in rotator cuff outcomes
One systematic review identified 13 trials involving 936 patients and reported pooled improvements in pain and disability after excluding studies judged at high risk of bias. The authors called for more trials and standardized protocols. These results support further investigation, while leaving uncertainty about durability and which patients benefit. They do not establish treatment of bone-on-bone arthritis. (9)
Another 2025 review found no clear pain or disability advantage
A different review included eight studies with 431 participants. It found no statistically significant advantage over controls for pain or shoulder disability. A small improvement in abduction was reported, averaging approximately seven degrees, while other movement measures were not significantly different. Changes in tendon imaging did not consistently correspond to clinical improvement. (10)
September 2026: no significant overall benefit across chronic shoulder pain trials
A review of 12 randomized trials found no statistically significant overall benefit for pain or function in its pooled analyses. An exploratory subgroup receiving at least three sessions appeared to have better pain results, but this was not a direct randomized comparison of one session against several. Differences between the studies were substantial. (11)
October 2026: rotator cuff and bursitis evidence remains uncertain
A newer review searched through March 2026 and included 14 studies; 10 were judged at high risk of bias. Most were small and short or medium term. For mixed rotator cuff tendinopathy and subacromial bursitis, the authors found that Prolotherapy may offer little or no benefit over saline, with uncertainty across other comparisons and limited adverse-event evidence. (12)
The reviews differ in study selection, comparison treatments, analysis, and assessment of bias. Taken together, they do not justify calling Prolotherapy a proven superior treatment for all shoulder disorders. They also do not erase an individual patient’s reported improvement. Research and patient accounts answer different questions.
Video case: advanced shoulder arthritis, early pain improvement, and continuing surgical candidacy
Danielle Matias, PA-C, presents a man in his mid-70s with advanced shoulder arthritis. The following is an edited transcript summary based on the supplied video transcript. It preserves the case findings and outcome without treating the presenter’s proposed mechanisms as established facts.
0:05–0:37: a golfer hoping to avoid replacement
The patient had experienced gradually increasing shoulder pain for approximately a year and a half. He could not identify a specific injury. He enjoyed golf and wanted less pain, better movement, and an alternative to shoulder replacement.
These were his goals. The video does not establish that he ultimately returned to golf or avoided surgery.
0:38–2:14: restricted movement and advanced joint changes
Danielle demonstrates motion X-rays taken while the patient raises his arm and reaches behind his body, similar to placing a hand in a back pocket. She describes severe joint-space narrowing, irregular bone surfaces, and bone spurs, calling the appearance bone on bone.
His active forward elevation was approximately 85 degrees, slightly below shoulder height. Abduction, lifting the arm out to the side, was approximately 70 degrees. Both were painful. Several ligament and tendon areas were markedly tender. Ultrasound was described as showing diffuse rotator cuff tendinopathy and cortical irregularity—an uneven appearance of the bone surface.
Dynamic imaging was useful to the presenter for illustrating movement and structural findings in this case. It does not directly measure pain or prove that every tender structure is a pain generator. An X-ray also does not directly display the cartilage; joint-space narrowing is an indirect sign of cartilage loss.
2:15–3:58: joint damage and soft-tissue contributors
Danielle considered the patient a shoulder replacement candidate. She also discussed the supporting soft tissues as possible contributors to pain.
Her explanation describes a clinical model in which incompletely healed ligament injury permits excessive small movements, or microinstability, which may alter joint loading. We retain that model as a possible contributor in selected shoulders. However, this patient did not recall an initiating injury, and the presentation does not establish that ligament laxity caused his arthritis. Shoulder osteoarthritis should not universally be described as beginning with a ligament injury.
3:59–4:41: Prolotherapy directed at multiple shoulder structures
The patient chose shoulder Prolotherapy. In describing this case, Danielle discusses treating the shoulder as a functional unit, with attention to supporting ligaments, rotator cuff structures, and the joint rather than the joint space alone.
This describes the approach presented in the video. It does not demonstrate that every injected structure regenerated or that a comprehensive injection approach is superior to other treatments for advanced arthritis.
4:42–5:03: early improvement in pain
At his second appointment, the patient reported that his pain had begun to decrease. It had not disappeared. Danielle explicitly states that the improvement did not change his status as a shoulder replacement candidate.
No numerical follow-up pain score, repeat range-of-motion measurement, or long-term outcome is supplied. The supported outcome is therefore early, patient-reported pain reduction.
5:04–6:02: what injections were not expected to change
Danielle emphasizes that Prolotherapy, PRP, and cell-based injections should not be presented as ways to remove established bone spurs or reverse the altered bone shape in this advanced shoulder. A mechanical restriction may remain even if pain improves.
Whether a bone spur needs surgical treatment depends on the clinical situation; the presence of a spur does not automatically require removal. Likewise, this case does not show that injections restored cartilage, recovered full movement, or eliminated a future need for replacement.
For a patient who cannot undergo surgery or strongly prefers to defer it, symptom improvement may still be meaningful. That possibility needs to remain separate from claims of structural restoration or a predictable treatment response.
Poor candidate for Prolotherapy
This x-ray shows an image of a patient that we would consider a poor candidate for Prolotherapy treatment. There is the loss of cartilage in the glenohumeral joint, and the patient has developed multiple bone spurs. These bone spurs significantly restricted his range of motion.
A patient who would be considered a fair candidate for Prolotherapy
In this X-ray, we see a patient with shoulder impingement and a moderately limited range of motion with abduction and external rotation. The X-ray image reveals a significant decrease in cartilage, which limited the patient’s shoulder and arm range of motion. This patient was considered a fair candidate for treatment, and realistic expectations of treatment outcomes should be discussed.
A patient who would be considered a good candidate for Prolotherapy
In this X-ray image, we see a patient who would be considered a good candidate for Prolotherapy. The overall architecture and integrity of the shoulder have been maintained in the joint. There is some loss of joint space. Because the overall shoulder structure has been maintained and the patient exhibited a full range of motion in his shoulder, this patient was considered a good candidate for Prolotherapy.
This presentation has fewer obvious structural barriers to movement. It still does not establish the source of pain or predict injection success. Research showing an imperfect relationship between radiographs and symptoms reinforces the need to interpret imaging in context. (2)
Structural findings and shoulder treatment considerations
This video discusses shoulder pain, arthritis, and structural considerations. It remains alongside the new case rather than being replaced.
We also discussed persistent pain after shoulder surgery and nerve-directed injections. Postoperative pain may involve several different problems; it should not automatically be assigned to the arthritic joint or treated as proof that replacement failed. Our shoulder replacement complications article retains that more specific discussion, including nerve-related pain and hydrodissection.
Additional Caring Medical research: shoulder labral lesions
Our published shoulder work also includes an uncontrolled questionnaire-based series of 33 patients with glenoid labral lesions. At a mean follow-up of 16 months, participants reported improvements in several symptoms and activities. One patient reported worsening of some symptoms. (13)
The labrum is the rim of tissue around the socket that contributes to stability. This series is relevant to our interest in shoulder supporting structures, but labral lesions are not the same diagnosis as advanced osteoarthritis. The study did not prove labral regrowth, provide a controlled comparison, or establish that Prolotherapy prevents replacement.
How PRP and other nonsurgical options fit
Prolotherapy, PRP, corticosteroids, and cell-based injections are different interventions. They should not be grouped together as though evidence for one proves effectiveness of all of them.
A randomized, double-blind trial assigned 70 patients with glenohumeral osteoarthritis to a single leukocyte-poor PRP or hyaluronic acid injection. Both groups improved, but there were no significant between-group differences in pain or functional outcomes through 12 months. Without a saline or no-injection group, the trial could not establish how much improvement was specific to either product. It did not demonstrate cartilage restoration. (14)
The 2020 AAOS guideline could not recommend injectable biologics such as PRP or stem cells for glenohumeral osteoarthritis because reliable evidence was lacking. That consensus statement predates the PRP trial above and is not a head-to-head finding about dextrose. It remains useful context for why symptom reports should not be described as established joint regeneration. (15)
Other options may include activity modification, exercise or physical therapy, appropriate pain medication, selected injections, and surgery. Their suitability depends on the condition and the person’s circumstances. Replacement remains an established option when arthritis causes substantial pain and functional loss despite nonsurgical care. Rotator cuff function and bone loss influence surgical choices. (15)
Our comparison of shoulder injections and discussion of shoulder replacement alternatives address those broader choices.
Questions this case and the research can answer
Can pain improve without the X-ray changing?
It can, but a reported improvement does not explain its own mechanism. Pain, movement, medication use, activity, and structural findings are separate outcomes. The golfer reported an early pain improvement; the presentation does not establish which tissues changed.
Can Prolotherapy remove a bone spur or guarantee a return to golf?
Those outcomes are not demonstrated here. The video explicitly acknowledges the structural limits, and it provides no confirmed return-to-golf outcome.
Does being a replacement candidate mean surgery is inevitable?
Candidacy means surgery may be a reasonable option; it does not establish what the patient ultimately chooses or needs. Conversely, an early response to injections does not prove that surgery will never be needed.
Are shoulder injections risk-free?
No. Potential complications include a pain flare, bleeding, infection, and injury to nearby structures. Small studies without serious reported events cannot exclude uncommon complications. The 2026 review found limited evidence about adverse events. (12)
Sudden severe shoulder pain after an injury, deformity, new marked weakness or numbness, or a hot swollen joint with fever requires prompt assessment. Shoulder or arm pain accompanied by chest pressure, breathlessness, sweating, or nausea may require emergency assessment rather than being assumed to be arthritis.
The clinical value of an improvement depends on what actually changed. Less pain may matter greatly, while movement restriction and advanced joint damage remain. Our aim in presenting this case and the research is to keep those distinctions clear.
Medical disclaimer: This article provides general education and does not establish an individual diagnosis or treatment recommendation. Patient accounts do not predict another person’s results. Evidence for Prolotherapy in advanced glenohumeral osteoarthritis remains limited; symptom improvement should not be interpreted as proven joint regeneration or avoidance of surgery.
References
1 Hauser RA, Lackner JB, Steilen-Matias D, Harris DK. A Systematic Review of Dextrose Prolotherapy for Chronic Musculoskeletal Pain. Clinical Medicine Insights: Arthritis and Musculoskeletal Disorders. 2016;9:139–159. doi:10.4137/CMAMD.S39160. [Google Scholar]
2 Joyce CD, Gutman MJ, Hill BW, Singh AM, Sherman M, Abboud JA, Namdari S. Radiographic Severity May Not be Associated with Pain and Function in Glenohumeral Arthritis. Clinical Orthopaedics and Related Research. 2022;480(2):354–363. doi:10.1097/CORR.0000000000001950. [Google Scholar]
3 Hauser RA, Hauser MA. A Retrospective Study on Hackett-Hemwall Dextrose Prolotherapy for Chronic Shoulder Pain at an Outpatient Charity Clinic in Rural Illinois. Journal of Prolotherapy. 2009;1(4):205–216. [Google Scholar]
4 Bertrand H, Reeves KD, Bennett CJ, Bicknell S, Cheng AL. Dextrose Prolotherapy Versus Control Injections in Painful Rotator Cuff Tendinopathy. Archives of Physical Medicine and Rehabilitation. 2016;97(1):17–25. doi:10.1016/j.apmr.2015.08.412. [Google Scholar]
5 Lee D H, Kwack KS, Rah UW, Yoon SH. Prolotherapy for Refractory Rotator Cuff Disease: Retrospective Case-Control Study of 1-Year Follow-Up. Archives of Physical Medicine and Rehabilitation. 2015;96(11):2027–2032. doi:10.1016/j.apmr.2015.07.011. [Google Scholar]
6 Seenauth C, Inouye V, Langland JO. Dextrose Prolotherapy for Chronic Shoulder Pain: A Case Report. Alternative Therapies in Health and Medicine. 2018;24(1):56–60. [Google Scholar]
7 Hsieh PC, Chiou HJ, Wang HK, Lai YC, Lin YH. Ultrasound-Guided Prolotherapy for Acromial Enthesopathy and Acromioclavicular Joint Arthropathy: A Single-Arm Prospective Study. Journal of Ultrasound in Medicine. 2019;38(3):605–612. doi:10.1002/jum.14727. [Google Scholar]
8 Robinson DM, Eng C, Makovitch S, Rothenberg JB, DeLuca S, Douglas S, Civitarese D, Borg-Stein J. Non-operative orthobiologic use for rotator cuff disorders and glenohumeral osteoarthritis: A systematic review. Journal of Back and Musculoskeletal Rehabilitation. 2021;34(1):17–32. doi:10.3233/BMR-201844. [Google Scholar]
9 Sony S, Kantha M, Shekhar S, Kumar A, Bhandari B, Muthuramalingam KP. Exploring dextrose prolotherapy in rotator cuff disorders: A systematic review and meta-analysis. Indian Journal of Anaesthesia. 2025;69(9):881–898. doi:10.4103/ija.ija_566_25. [Google Scholar]
10 Thamrongskulsiri N, Vitoonpong T, Itthipanichpong T, Limskul D, Tanpowpong T, Kuptniratsaikul S. Prolotherapy is not superior to control or placebo-based conservative treatments for rotator cuff tendinopathy: a systematic review and meta-analysis. Clinics in Shoulder and Elbow. 2025;28(4):446–456. doi:10.5397/cise.2025.00570. [Google Scholar]
11 Darmawan R, Waluyo Y, Sam N, Zainal ATF, Makkaraka MAG, Go R. The effect of dextrose prolotherapy on pain and functional outcomes in chronic shoulder pain: A systematic review and meta-analysis of randomized controlled trials. Qatar Medical Journal. 2026;2026(3):62. doi:10.5339/qmj.2026.62. [Google Scholar]
12 Zerzan N, Ewart D, Sowerby C, Yang S, Kalinowski C, Ullman K, Landsteiner A, Duan-Porter W, Wilt TJ. Dextrose prolotherapy for chronic symptomatic rotator cuff tendinopathy and subacromial bursitis: A systematic review and meta-analysis. PM&R. Published online October 2, 2026. doi:10.1002/pmrj.70219. [Google Scholar]
13 Hauser RA, Dolan E, Orlofsky A. Prolotherapy: A Non-Invasive Approach to Lesions of the Glenoid Labrum; A Non-Controlled Questionnaire Based Study. The Open Rehabilitation Journal. 2013;6:69–76. doi:10.2174/1874943701306010069. [Google Scholar]
14 Kirschner JS, Cheng J, Creighton A, Santiago K, Hurwitz N, Dundas M, Beatty N, Kingsbury D, Konin G, Abutalib Z, Chang R. Efficacy of Ultrasound-Guided Glenohumeral Joint Injections of Leukocyte-Poor Platelet-Rich Plasma Versus Hyaluronic Acid in the Treatment of Glenohumeral Osteoarthritis: A Randomized, Double-Blind Controlled Trial. Clinical Journal of Sport Medicine. 2022;32(6):558–566. doi:10.1097/JSM.0000000000001029. [Google Scholar]
15 Khazzam MS, Pearl ML. AAOS Clinical Practice Guideline: Management of Glenohumeral Joint Osteoarthritis. Journal of the American Academy of Orthopaedic Surgeons. 2020;28(19):790–794. doi:10.5435/JAAOS-D-20-00405. [Google Scholar]