Best evidence-based regenerative medicine therapy for chronic joint pain.
Judged strictly on proper trials, the best therapies for long-term joint pain are exercise, weight loss (including semaglutide for people with obesity), anti-inflammatory gels, and — for pain that has spread — duloxetine. The only regenerative therapy with strong trial evidence is cartilage cell implantation (MACI) for a specific cartilage defect. PRP helps a little and inconsistently; stem-cell injections have not beaten placebo in the best trials; exosomes have no trials. Ask any clinic to name the trial behind what it sells.
Ranked strictly on randomised trials with the joint, the population and the outcome named, the best evidence-based therapies for chronic joint pain are: structured exercise (dozens of trials in knee and hip osteoarthritis; pain and function; durable); weight loss, with semaglutide in obesity (a 68-week randomised trial: large reduction in knee pain versus placebo); topical NSAIDs (randomised trials in knee and hand osteoarthritis); autologous chondrocyte implantation (randomised against microfracture for focal cartilage defects — the one regenerative therapy with an A); duloxetine (randomised trials in osteoarthritis pain with central sensitisation); corticosteroid injection (short-term relief in trials; cartilage loss with repetition in a two-year trial); PRP (meta-analyses of knee osteoarthritis trials: modest benefit over hyaluronic acid, inconsistent over placebo; the largest placebo-controlled trial null on pain and MRI); hyaluronic acid (many trials; small effect judged not clinically important by several guidelines); mesenchymal cell injections (randomised trials inconsistent; better-controlled ones null; no approval); exosomes (no trials). 'Evidence-based regenerative therapy for chronic joint pain' therefore names exactly one therapy for one indication — MACI for a cartilage defect — and a modest adjunct, PRP, for the rest. The evidence-based treatment of chronic joint pain is mostly not regenerative, and it works.
- Evidence-based means a randomised trial in this joint with this outcome; every line here names one, and the marketing usually does not.
- The strongest trial evidence for chronic joint pain is for exercise, weight loss and semaglutide — none regenerative.
- MACI is the only regenerative therapy with randomised evidence and an approval, for a focal cartilage defect.
- PRP's evidence is real, modest and inconsistent; cell injections' evidence is inconsistent and, where best-controlled, null.
- A therapy sold as evidence-based should be able to name its trial; if it cannot, it is not.
Chronic joint pain therapies ranked on the trials
Ranked on: the quality, size, population and outcome of the randomised trials for each therapy in chronic joint pain — osteoarthritis unless stated — with the trial evidence described on every line. A therapy without a randomised trial in a joint cannot rank above D.
| # | Option | Verdict | Grade |
|---|---|---|---|
| 1 | Structured exercise therapy | Dozens of RCTs; pain and function; durable | GRADE AEstablished |
| 2 | Weight loss, and semaglutide in obesity | A 68-week placebo-controlled RCT; large pain reduction | GRADE AEstablished |
| 3 | Topical NSAIDs | RCTs in knee and hand osteoarthritis; low systemic risk | GRADE AEstablished |
| 4 | Autologous chondrocyte implantation (MACI) — focal cartilage defect | RCT versus microfracture; approved; the regenerative A | GRADE AEstablished |
| 5 | Duloxetine — osteoarthritis pain with central sensitisation | Randomised trials; pain reduction; a defined subgroup | GRADE BPromising |
| 6 | Corticosteroid injection | RCTs: short-term relief; a two-year RCT: cartilage loss | GRADE BPromising |
| 7 | PRP injection — knee osteoarthritis | Meta-analyses: modest vs HA, inconsistent vs placebo; largest trial null | GRADE CEarly |
| 8 | Hyaluronic-acid injection | Many RCTs; small effect; judged not clinically important | GRADE CEarly |
| 9 | Mesenchymal cell injections — bone marrow, adipose, cultured | RCTs inconsistent; best-controlled null; no approval | GRADE CEarly |
| 10 | Prolotherapy | Small RCTs; small effects | GRADE CEarly |
| 11 | Glucosamine and chondroitin | Large independent RCT null | GRADE DInsufficient or unsafe |
| 12 | Exosomes, IV cell products, 'regenerative' infusions | No randomised trial in any joint | GRADE DInsufficient or unsafe |
- 01
Structured exercise therapy
GRADE AEstablishedDozens of RCTs; pain and function; durableSystematic reviews of randomised trials in knee and hip osteoarthritis — land-based, aquatic, strengthening, aerobic — show consistent moderate reductions in pain and improvements in function that persist after supervised programmes end. The largest and most replicated evidence base for any chronic-joint-pain therapy.
- 02
Weight loss, and semaglutide in obesity
GRADE AEstablishedA 68-week placebo-controlled RCT; large pain reductionWeight-loss trials show pain reduction proportional to weight lost in knee osteoarthritis. In a randomised, placebo-controlled trial of weekly semaglutide in adults with obesity and knee osteoarthritis over 68 weeks, weight fell substantially more and knee pain scores fell markedly more than with placebo. Named trial, named population, named outcome.
- 03
Topical NSAIDs
GRADE AEstablishedRCTs in knee and hand osteoarthritis; low systemic riskRandomised trials of topical diclofenac and related agents show pain relief in knee and hand osteoarthritis comparable to oral NSAIDs over weeks with minimal systemic exposure. First-line in guidelines on that evidence.
- 04
Autologous chondrocyte implantation (MACI) — focal cartilage defect
GRADE AEstablishedRCT versus microfracture; approved; the regenerative AA randomised trial comparing MACI with microfracture for symptomatic full-thickness cartilage defects of the knee showed better pain and function at two years, sustained at five. Approved on that evidence. The only regenerative therapy for a joint with a randomised trial, a positive result and an approval — for a defect, not for osteoarthritis.
- 05
Duloxetine — osteoarthritis pain with central sensitisation
GRADE BPromisingRandomised trials; pain reduction; a defined subgroupRandomised, placebo-controlled trials in knee osteoarthritis show duloxetine reduces pain and improves function, with larger effects where central sensitisation is present. Nausea and blood-pressure effects; serotonergic interactions. Evidence-based for the pain the joint does not explain.
- 06
Corticosteroid injection
GRADE BPromisingRCTs: short-term relief; a two-year RCT: cartilage lossRandomised trials show pain relief over weeks in osteoarthritis flares. A two-year randomised trial of triamcinolone every three months versus saline found greater cartilage volume loss with no lasting pain benefit. Evidence-based as an occasional bridge; evidence-against as a schedule.
- 07
PRP injection — knee osteoarthritis
GRADE CEarlyMeta-analyses: modest vs HA, inconsistent vs placebo; largest trial nullMeta-analyses of randomised trials show PRP outperforming hyaluronic acid on pain at six to twelve months and inconsistently outperforming placebo; the largest placebo-controlled trial found no difference in pain or cartilage volume on MRI at twelve months. Preparations vary. Evidence-based as a modest, uncertain adjunct.
- 08
Hyaluronic-acid injection
GRADE CEarlyMany RCTs; small effect; judged not clinically importantNumerous randomised trials show a small benefit over placebo lasting weeks to months that several guideline bodies judge below the threshold of clinical importance. Evidence exists; the effect is small.
- 09
Mesenchymal cell injections — bone marrow, adipose, cultured
GRADE CEarlyRCTs inconsistent; best-controlled null; no approvalRandomised trials of MSC injection for knee osteoarthritis report mixed results, with the better-controlled trials showing no advantage over placebo or hyaluronic acid on pain or cartilage; no product has approval for osteoarthritis. Evidence exists and does not support use outside a trial.
- 10
Prolotherapy
GRADE CEarlySmall RCTs; small effectsSmall randomised trials in knee osteoarthritis suggest modest benefit over saline or exercise alone; heterogeneous protocols. Low risk; weak evidence.
- 11
Glucosamine and chondroitin
GRADE DInsufficient or unsafeLarge independent RCT nullA large independent randomised trial found glucosamine, chondroitin and the combination no better than placebo for knee osteoarthritis pain overall; industry-funded trials are more positive. Evidence-based to skip.
- 12
Exosomes, IV cell products, 'regenerative' infusions
GRADE DInsufficient or unsafeNo randomised trial in any jointNo trial evidence for chronic joint pain of any kind; unverified products; documented harms from unapproved cell products. Not evidence-based in any sense of the phrase.
Name the trial
The evidence behind each therapy, stated
| Therapy | Best trial evidence | Population | Outcome | Result |
|---|---|---|---|---|
| Exercise | Dozens of RCTs; systematic reviews | Knee and hip OA | Pain, function | Moderate benefit, durable |
| Semaglutide | 68-week placebo-controlled RCT | Obesity + knee OA | Weight, knee pain | Large benefit |
| Topical NSAIDs | RCTs; systematic reviews | Knee and hand OA | Pain | Benefit comparable to oral |
| MACI | RCT vs microfracture, 2–5 years | Focal cartilage defect | Pain, function | Superior; approved |
| Duloxetine | Placebo-controlled RCTs | Knee OA | Pain, function | Benefit; larger with sensitisation |
| Corticosteroid | RCTs; 2-year RCT vs saline | Knee OA | Pain; cartilage volume | Short relief; more cartilage loss |
| PRP | Meta-analyses; large placebo RCT | Knee OA | Pain; MRI | Modest, inconsistent; largest trial null |
| Hyaluronic acid | Many RCTs | Knee OA | Pain | Small; below clinical importance |
| MSC injection | Several RCTs | Knee OA | Pain; cartilage | Inconsistent; best-controlled null |
| Glucosamine/chondroitin | Large independent RCT | Knee OA | Pain | Null |
| Exosomes | None | — | — | — |
Frequently asked questions
What is the best evidence-based regenerative therapy for chronic joint pain?
Strictly on randomised trials: autologous chondrocyte implantation (MACI) for a focal cartilage defect — randomised against microfracture, approved — is the only regenerative therapy with an A. PRP is a modest, inconsistent adjunct for osteoarthritis; cell injections are null in the best-controlled trials; exosomes have no trials. The best evidence-based therapies for chronic joint pain overall are exercise, weight loss with semaglutide in obesity, topical NSAIDs and duloxetine — none regenerative.
What does 'evidence-based' actually require?
A randomised controlled trial in the joint being treated, in patients like you, measuring the outcome you care about — pain, function or structure — with a result that favours the therapy. Every line in this guide names its trial evidence. A therapy whose seller cannot name a trial meeting that description is not evidence-based, whatever the brochure says.
Is PRP evidence-based for knee osteoarthritis?
Partly. Meta-analyses of randomised trials show PRP outperforms hyaluronic acid on pain at six to twelve months and inconsistently outperforms placebo; the largest placebo-controlled trial found no difference in pain or cartilage on MRI at one year. The evidence supports it as a modest, uncertain adjunct once exercise and weight are addressed — not as a primary therapy.
What do the stem cell trials for joint pain actually show?
Randomised trials of mesenchymal cell injection for knee osteoarthritis are inconsistent, and the better-controlled ones show no advantage over placebo or hyaluronic acid on pain or cartilage; no product is approved for osteoarthritis. The evidence exists, and it does not support use outside a clinical trial.
Which non-regenerative therapies have the best trial evidence for joint pain?
Exercise (dozens of randomised trials in knee and hip osteoarthritis, durable benefit), weight loss with semaglutide in obesity (a 68-week placebo-controlled trial with a large reduction in knee pain), topical NSAIDs (randomised trials in knee and hand osteoarthritis) and duloxetine (placebo-controlled trials, particularly with central sensitisation). These are the therapies to build a chronic-joint-pain plan around.
Is MACI available for ordinary arthritis?
No. Its randomised evidence and approval are for symptomatic full-thickness cartilage defects — usually after injury in a younger patient — where it outperformed microfracture at two to five years. It does not treat the diffuse cartilage loss of osteoarthritis, and it is a surgical procedure from an orthopaedic surgeon, not an injection from a clinic.
Keep reading
- Which regenerative medicine treatments work best for joint pain?
The same therapies with a pharmacist's sequence.
- Regenerative medicine: what is approved, what is in trials, and what is only being sold
The full evidence ledger with trial identifiers.
- How we grade evidence
Why a null trial still earns a C and no trial earns a D.
- Free stack check
Screen duloxetine and NSAIDs against everything else you take.
More in Regenerative medicine
- Which regenerative medicine treatments work best for joint pain?
Regenerative treatments for joint pain ranked against what actually works: exercise therapy and weight loss, GLP-1 agonists for knee OA, corticosteroid and hyaluronic-acid injections, PRP, bone-marrow and adipose stem-cell injections, autologous chondrocyte implantation, exosomes and clinic 'stem cells' — with a pharmacist's advice on what to try first.
- What is the most effective regenerative medicine for knees?
Knee treatments ranked on randomised evidence — exercise, weight loss and semaglutide, MACI for cartilage defects, steroid, PRP, hyaluronic acid, bone-marrow and adipose cell injections, exosomes — separated by what is wrong with the knee: osteoarthritis, a cartilage defect, a meniscal tear or a ligament injury.
- Which regenerative medicine options help avoid joint replacement?
Options for delaying or avoiding joint replacement ranked on evidence: exercise and weight loss, GLP-1 agonists, bracing and offloading, osteotomy, MACI for cartilage defects, steroid and PRP injections, stem-cell injections, exosomes — and when delaying a replacement does more harm than the operation.
- What regenerative medicine is best for chronic back pain?
Chronic back pain treatments ranked on randomised evidence: exercise and cognitive-behavioural approaches, multidisciplinary rehabilitation, staying active, epidural and facet injections, radiofrequency ablation, intradiscal PRP and cell injections, basivertebral nerve ablation, spinal fusion — and the regenerative injections that have not beaten placebo.
- Which regenerative medicine therapy is safest for arthritis patients?
Arthritis therapies ranked on safety for the arthritis patient: exercise and weight loss, topical NSAIDs, PRP, hyaluronic acid, corticosteroid injections, MACI, GLP-1 agonists, oral NSAIDs, bone-marrow and adipose cell injections, exosomes and clinic 'stem cells' — with the documented harms, the interactions and what an inflammatory-arthritis patient must know.
- Best regenerative medicine treatments for osteoarthritis relief without surgery.
Non-surgical osteoarthritis treatments ranked on relief and durability: exercise, weight loss and semaglutide, topical NSAIDs, braces and aids, steroid injections, PRP, hyaluronic acid, duloxetine, stem-cell injections and exosomes — assembled into a twelve-week programme a pharmacist would prescribe.