Muscle, joint & pain

Prolotherapy for Joints and Tendons, Read Against the Data

Save

An injection that has been offered for decades, on a theory that sounds almost plausible: irritate the tissue a little, and the body will repair it. The interesting question is not whether the theory is elegant. It is where prolotherapy sits on the map that clinical guidelines draw — the map with strong on one end, recommended against on the other, and a great deal of silence in between.

Last updated: July 2026

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

What is prolotherapy?

An injection of an irritant, given as a course rather than a one-off. The solution is most often concentrated dextrose — sugar water, essentially — placed into a joint, a ligament, or the point where a tendon anchors to bone. The proposed mechanism is deliberate provocation: irritate the tissue enough to summon a healing response the body did not mount on its own.

Prolotherapy — from proliferation therapy — is the oldest of the injection treatments that promise repair rather than relief. It predates platelet-rich plasma by decades, and that is worth holding onto: it means the treatment has had a long time to accumulate evidence, and a long time for that evidence to be examined.

The theory is not absurd. Inflammation is part of how tissue heals, and the idea that a controlled irritation might restart a stalled repair is coherent enough to be worth testing. Coherent and true are different properties, though, and the distance between them is the thing trials exist to measure.

Why are so many people shopping for an injection?

Because the underlying problem is enormous and the settled treatments are slow. Roughly 58.5 million US adults — about 23.7 percent — had doctor-diagnosed arthritis in 2016 to 2018, and around 25.7 million reported that arthritis limited what they could do 1. That is a very large number of people who have already tried the patient thing and want something else.

About 25.7 million US adults report that arthritis limits their activity 1. Knee arthritis alone comes in several forms — osteoarthritis, rheumatoid, post-traumatic — and the menu for it runs from nonsurgical measures through to an operation 2.

This is context, not cynicism. Demand at that scale attracts careful clinicians and confident ones alike, and it means a person researching prolotherapy at midnight is not being foolish. They are doing what people do when the honest answer they were handed was months of exercise and the pain is now.

How to place any therapy on the guideline map

By finding its grade and reading the grade's actual wording. Clinical guidelines do not simply list treatments; they rank them by how good the evidence is. The ACR and Arthritis Foundation guideline for hand, hip, and knee osteoarthritis strongly recommends exercise, weight loss for people carrying extra weight, and self-management, and conditionally recommends topical and oral NSAIDs and corticosteroid injections into the joint 3.

Four positions exist on that map, and they are not the same thing:

  • Strongly recommended. The panel judged the evidence good and the balance of benefit and harm clear. The AAOS guideline for knee osteoarthritis places exercise and physical therapy, NSAIDs, and weight loss in this category 4.
  • Conditionally recommended. Worth offering, with reservations — usually because the benefit is modest, or short, or uncertain. Corticosteroid injection sits here 3.
  • Recommended against. The panel looked and decided the evidence does not support it. The ACR guideline does this to several therapies it judged low-value 3.
  • Absent. The therapy is not in the document at all.

That last position is the one that matters here, and it is the one people misread. This page will not tell you which position prolotherapy occupies, because the sources behind this page do not name it, and assigning it a grade from memory would be precisely the failure the page is arguing against. The move is to open the guideline for your joint and look. The same exercise works on anything else a clinician offers you: an epidural steroid injection for sciatica, for instance, has its own trial record and its own grade.

Guidelines are willing to say no

Which is why absence is informative rather than suspicious. Guideline panels are not shy about recommending against a popular procedure. The BMJ Rapid Recommendation on arthroscopic surgery for degenerative knee disease issued a strong recommendation against it for nearly all patients — including those with meniscal tears, mechanical symptoms, or sudden onset 5. That is a panel telling surgeons to stop doing an operation, which is not the behaviour of a body protecting the status quo.

The familiar story about unconventional treatments is that the establishment ignores them out of self-interest. The record does not support the story's premise. The same guideline literature recommends against a surgery its own members perform 5, and against several therapies it judged low-value 3. A panel willing to say no to arthroscopy would have no particular difficulty saying no to sugar water.

Absence from a guideline usually means the evidence was too thin to grade, not that the therapy was suppressed. Thin evidence is not disproof — that is genuinely a different finding, and it is the honest thing to say about a treatment nobody has tested well. But thin evidence is not a green light either, and the cost of a course is real whichever way the evidence eventually falls.

Why enthusiastic early results are not enough

Because a series of injections judged at twelve weeks can tell you very little about a year. The clearest published demonstration is not about prolotherapy at all — it is about cortisone at the elbow. In a randomized trial, people with tennis elbow who received a corticosteroid injection were worse at one year, with more recurrence, than people who received a placebo injection 6.

Two features of that trial make it useful here, and neither is the headline. The first is the placebo arm: without a group receiving a needle full of nothing, the trial could never have discovered that the needle full of something was the worse option. The second is the calendar. Stop the study early and you publish a different paper.

Then there is the plain fact that people book an injection at their worst. Pain that has spiked is pain with room to fall on its own, and any treatment delivered at the peak inherits the credit for the descent. This is why a grateful series of patients — which is what a testimonial is — cannot distinguish a working treatment from a calendar.

The regenerative-injection family

Prolotherapy has relatives, and they are sold in much the same register. Platelet-rich plasma spins your own blood and reinjects the platelet fraction. Stem-cell preparations go further, and the claims go further with them. Hyaluronic acid injections are a slightly different animal — a lubricant story rather than a repair story — but they compete for the same appointment and the same money.

Each has its own trials, its own grade, and its own page in this library: prp evidence for the platelet preparations, stem cell injection evidence for the cell-based ones, hyaluronic acid injection for the gels, and prp for knee arthritis for the specific knee question. They are worth reading separately, because the family resemblance lives in the pitch rather than in the data. Nothing about one of them predicts another.

What they do share is a shape: a mechanism described in the language of healing rather than relief, and a course of appointments rather than a single visit. That shape is not evidence of anything, in either direction. It is simply what is being offered, and it is worth naming out loud before the first appointment rather than after the fourth.

What the strong grades landed on, and where a knife still fits

On the unglamorous things. Across the osteoarthritis guidelines the pattern is the same: exercise, weight loss for people carrying extra weight, and self-management carry the strong recommendations 3, and the AAOS guideline grades exercise and physical therapy, NSAIDs, and weight loss as strong evidence for the knee 4. In neither document does an injection appear at that tier.

None of which is an argument against intervening, and this page would be misread as one at the reader's cost. The same ACR document conditionally recommends corticosteroid injection into the joint 3 — a real endorsement, on narrow terms. Orthopaedic patient guidance for knee arthritis presents surgery as one of the genuine options rather than a failure state, with the choice turning on the specifics of the joint and the person 2. And the BMJ recommendation is against one operation for one condition; its own wording is nearly all patients, not all 5. Panels write nearly for a reason.

The question is never whether to intervene. It is what has been tried, in what order, and what a given intervention is supposed to buy. This page cannot hand you the list of situations where an operation is clearly the right call, because that list is joint-specific and these sources are about managing osteoarthritis rather than about surgical indications. What it can say is that the list is real, that it belongs to the guideline for your joint and to a surgeon who has examined it, and that nothing above is a reason for delay once a joint has reached that point. Whether knee replacement is worth it, for one, is a separate question with a separate literature and a page of its own.

The modest proposal underneath all of this: before buying a course of injections nobody has graded, it is worth knowing whether the graded things have actually been done.

Common questions

The sources behind this page do not grade prolotherapy, so answering either way here would be dishonest. What can be said is where to find out. The clinical guideline for the joint in question states its recommendations by grade, and a therapy is either strongly recommended, conditionally recommended, recommended against, or absent. Absent usually means the evidence was too thin to grade.

No. Prolotherapy injects an irritant solution, usually concentrated dextrose, to provoke a repair response. Platelet-rich plasma spins your own blood and reinjects the concentrated platelet fraction on a growth-factor rationale. They share a marketing category and a promise of healing rather than relief, but they are different substances with separate trial records, and evidence about one says nothing about the other.

The record makes that hard to argue. The same guideline literature issued a strong recommendation against arthroscopic surgery for degenerative knee disease in nearly all patients, and the ACR guideline recommends against several therapies it judged low-value. Panels willing to tell surgeons to stop operating are not obviously protecting anyone. Absence far more often reflects evidence too thin to grade than evidence that was buried.

Because people book an injection when the pain is at its worst, and pain at its worst has room to fall on its own. Any treatment given at the peak collects credit for the descent. A testimonial cannot separate the injection from the calendar, from the rest, or from the attention. Only a group given a dummy injection can do that, which is what a placebo arm is for.

Exercise, weight loss for people carrying extra weight, and self-management. The ACR and Arthritis Foundation guideline puts those in its strong tier for hand, hip, and knee osteoarthritis, and the AAOS guideline grades exercise and physical therapy, NSAIDs, and weight loss as strong evidence for the knee. No injection appears at that tier in either document. That is the comparison any injection has to beat.

Related

Say it back

How would you explain this to someone you love?

Two or three sentences, just as you’d say it. Gale reflects back what you focused on — a mirror, not a quiz.

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

When a joint has stopped being an injection question

  • —A knee that locks in a bent position and cannot be straightened, or a joint that gives way and drops you.
  • —A joint that is hot, red, and swollen along with a fever or a general feeling of being unwell — particularly within days of any injection into it.
  • —A sudden inability to put weight through the limb after a fall, or a limb that looks deformed or out of position.
  • —Numbness, weakness, or pins and needles in the limb, or pain that wakes you at night and does not settle with a change of position.

A hot, swollen, painful joint together with a fever — above all in the days after an injection into that joint — belongs in an emergency department the same day. A joint infection is treated in hours, not at the next available appointment.

Gale's library explains what the published evidence says and how to read it. It has not examined your joint and does not know your history. Nothing here is a treatment recommendation for you, and what belongs in your own sequence of care is a conversation with a clinician who can assess the joint itself.

References

  1. 1.Theis KA, Murphy LB, Guglielmo D, et al. (CDC/MMWR) (2021). Prevalence of Arthritis and Arthritis-Attributable Activity Limitation — United States, 2016-2018. MMWR (CDC Morbidity and Mortality Weekly Report). linkAbout 58.5 million US adults (23.7 percent) had doctor-diagnosed arthritis in 2016-2018, with roughly 25.7 million reporting arthritis-attributable activity limitation — the article's account of the scale of demand that an unproven injection is being offered into.
  2. 2.American Academy of Orthopaedic Surgeons (OrthoInfo) (2024). Arthritis of the Knee. OrthoInfo — AAOS. link ✓Lay-education framing that knee arthritis comes in several forms (osteoarthritis, rheumatoid, post-traumatic) and that its treatment menu runs from nonsurgical measures through to surgical options — used for the article's claim that orthopaedic guidance presents surgery as a genuine option rather than a failure state.
  3. 3.Kolasinski SL, Neogi T, Hochberg MC, et al. (2020). 2019 American College of Rheumatology/Arthritis Foundation Guideline for the Management of Osteoarthritis of the Hand, Hip, and Knee. Arthritis & Rheumatology / Arthritis Care & Research. doi:10.1002/art.41142 ✓The guideline strongly recommends exercise, weight loss for people who are overweight, and self-management; conditionally recommends topical/oral NSAIDs and intra-articular corticosteroids; and recommends against several low-value therapies — the article's worked example of guideline grading language and of panels being willing to recommend against a treatment.
  4. 4.American Academy of Orthopaedic Surgeons (AAOS) (2021). Management of Osteoarthritis of the Knee (Non-Arthroplasty), Third Edition — Clinical Practice Guideline. AAOS. link ✓The 2021 third-edition guideline grades exercise and physical therapy, NSAIDs, and weight loss as supported by strong evidence for non-arthroplasty management of knee osteoarthritis — the article's illustration of what the strong tier contains.
  5. 5.Siemieniuk RAC, Harris IA, Agoritsas T, et al. (2017). Arthroscopic surgery for degenerative knee arthritis and meniscal tears: a clinical practice guideline. BMJ. doi:10.1136/bmj.j1982A strong recommendation against arthroscopy for nearly all patients with degenerative knee disease, including those with meniscal tears, mechanical symptoms, or acute onset — the article's evidence that guideline panels will recommend against a popular procedure, and that the wording 'nearly all' is deliberate.
  6. 6.Coombes BK, Bisset L, Brooks P, Khan A, Vicenzino B (2013). Effect of Corticosteroid Injection, Physiotherapy, or Both on Clinical Outcomes in Patients With Unilateral Lateral Epicondylalgia: A Randomized Controlled Trial. JAMA. PMID 23385272 ✓For tennis elbow, corticosteroid injection produced worse outcomes and higher recurrence at one year than placebo injection — used here as the methodological demonstration that an injection can look good early and lose to placebo late, and that the placebo arm and the follow-up window are what make a trial informative.

6 sources, numbered by first appearance. General health information, not medical advice. AI-assisted editorial content — citations link their sources. Editorial policy

← Back to the health library