Red Light Therapy for Knee Osteoarthritis: What the Clinical Trials Actually Show

Knee osteoarthritis wears down the cartilage that cushions the joint, and it affects hundreds of millions of people. Red light therapy, also called photobiomodulation or low-level laser therapy (LLLT), is one of the more controversial treatments aimed at easing that pain. The research is real but messy: some high-quality trials show meaningful relief, others show nothing better than a placebo, and the disagreement often comes down to one boring detail nobody talks about in marketing copy.

This review walks through what the clinical trials actually found, why the meta-analyses contradict each other, where the major medical guidelines land, and how to read a study before you spend money. It is written to be honest about the gaps, not to sell you a panel.

What "Red Light Therapy" Means for an Arthritic Knee

The terms get used loosely, so it helps to be precise. The same core idea shows up under several names:

• Photobiomodulation (PBM) is the modern umbrella term researchers prefer.
• Low-level laser therapy (LLLT) uses a laser diode, usually delivered point by point over the knee.
• LED red light therapy uses light-emitting diodes instead of lasers, the kind in consumer panels and wraps.
• High-intensity laser therapy (HILT) uses much higher power and is a separate, related treatment.

For knee osteoarthritis, the bulk of the published trials use low-level lasers, not the LED panels most people buy for home use. That distinction matters when you try to apply trial results to a device on your shelf. A 904 nm laser delivering a measured dose to nine spots around the joint is not the same as standing in front of a full-body panel.

The Proposed Mechanism

The cellular theory is reasonably well established even where the clinical results are not. Red and near-infrared light in the roughly 600 to 1,000 nm range is absorbed by cytochrome c oxidase, an enzyme in the mitochondria. That absorption is thought to nudge cells to produce more ATP (cellular energy), shift signaling molecules like nitric oxide, and dampen inflammatory mediators.

In an arthritic knee, the hoped-for downstream effects are less inflammation in the joint lining, less pain signaling, and possibly some protection of cartilage. The mechanism is plausible and supported by lab work. But a plausible mechanism is not proof of clinical benefit, and that gap is the whole story here. We cover the cellular side in more depth in our photobiomodulation science guide.

Why the Knee Is a Hard Target

There is a physics problem worth naming early. Light loses intensity fast as it travels through tissue, and the knee joint sits beneath skin, fat, fascia, and the joint capsule. Near-infrared wavelengths around 800 to 900 nm penetrate deeper than visible red, which is part of why so many knee trials use them. But even near-infrared light scatters and weakens with every millimeter of tissue. By the time it reaches the synovium or the cartilage surface, only a fraction of the surface dose remains.

That falloff is exactly why dose at the skin matters so much, and why a weak device or a short session may deliver nothing useful to the actual joint. A clinic that presses a laser probe directly against the knee at several points is trying to push enough energy through that tissue stack to matter. A panel sitting a foot away, radiating a large body area at low intensity, faces a much steeper uphill climb to deliver a therapeutic dose to a structure that deep. None of the trials have settled the ideal way to get past this barrier, which is one more reason the literature is so inconsistent.

What the Clinical Trials Actually Show

Here is the honest summary: the evidence is mixed, and the two largest meta-analyses reach opposite conclusions. That is unusual, and it is the single most important thing to understand before believing any headline.

The Negative Meta-Analysis (2015)

A systematic review published in Osteoarthritis and Cartilage pooled nine randomized controlled trials covering 518 patients. The result was blunt. There was no significant difference between laser therapy and placebo for pain right after treatment (standardized mean difference of −0.28, with a confidence interval crossing zero), and no significant difference at the 12-week follow-up. Even the trials that followed the World Association of Laser Therapy (WALT) dosing recommendations showed no clear benefit. The authors concluded the best available evidence at the time did not support LLLT for knee osteoarthritis (Huang et al., 2015, PMID 25914044).

The Positive Meta-Analysis (2019)

Four years later, a larger analysis in BMJ Open pooled 22 randomized placebo-controlled trials with 1,063 patients and reached a very different conclusion, with one crucial twist: dose changed everything.

When the researchers separated trials by whether they used the WALT-recommended laser dose, a pattern appeared that the 2015 review had washed out by lumping all doses together (Stausholm et al., 2019, PMID 31662383).

A 20 to 30 mm improvement on a 100 mm pain scale is clinically meaningful, not a rounding error. The catch is the dose ranges the authors flagged as effective: 4 to 8 joules per spot at 785 to 860 nm, or 1 to 3 joules per spot at 904 nm. Outside those windows, the benefit largely disappeared.

So the two meta-analyses do not actually contradict each other as much as it first looks. The 2015 review pooled effective and ineffective doses together and got a wash. The 2019 review separated them and found the dose is the lever.

This is a recurring trap in light therapy research. When you average a group of trials where half used a real, effective dose and half used too little energy, the strong results and the null results cancel out, and the pooled number lands near zero. A reader skimming the abstract walks away thinking "it does not work," when the honest conclusion is "it does not work at the wrong dose." The 2019 team's contribution was less about new patients and more about refusing to blend those two populations. Their subgroup test for the difference between recommended and non-recommended doses was statistically significant at both the immediate and follow-up timepoints, which is the kind of finding that should make anyone reading older laser reviews go back and check what doses were actually pooled.

The Quality Problem Nobody Advertises

Even the positive findings come with caveats that matter. Many of the included trials were small, enrolling a few dozen patients each. Blinding was inconsistent, and a fake laser is genuinely hard to blind well because patients and operators sometimes sense which device is "on." Treatment protocols varied wildly in number of sessions, spots treated, and total energy delivered. Outcome scales differed. When formal grading systems are applied to this body of work, the certainty of evidence usually lands at low to very low, even for the outcomes that look positive.

That does not erase the positive signal. It means the signal is real but fragile, and a few large, well-blinded trials could shift the picture in either direction. Anyone who tells you the question is settled, in either direction, is overstating what 22 mostly small trials can prove.

A Recent Single Trial (2025)

A 2025 double-blind randomized controlled trial added a fresh data point. It enrolled 73 patients with symptomatic knee osteoarthritis, splitting them into photobiomodulation, sham, and no-treatment groups. The active group received 790 nm light at 4 joules per point across nine areas of the knee and showed significantly greater pain reduction than both the sham and control groups (PMID 40545487). It is one trial, not a body of evidence, but it lines up with the dose-dependent story.

Laser Plus Exercise

Several reviews looked at laser therapy added on top of an exercise program rather than alone. A 2023 systematic review and meta-analysis found that LLLT combined with exercise significantly reduced pain compared with exercise alone, though it found no significant added benefit for range of motion, muscle strength, or function (PMID 36576096). A 2024 network meta-analysis comparing both low-level and high-intensity laser as add-ons to exercise found both beat placebo-plus-exercise on pain and WOMAC function scores at four and eight weeks (PMID 39193110).

The takeaway is consistent across these: laser looks best as a complement to exercise, not a replacement for it. Exercise remains the strongly recommended foundation; light is, at best, an add-on.

One honest worry about the laser-plus-exercise trials: it is hard to know how much credit belongs to the light and how much to the exercise. If both groups exercise and only one gets laser, a difference points to the light. But if adherence to the home exercise program differed between groups, or if the laser sessions simply got people into the clinic more often, some of the measured benefit may be doing-something effects rather than photobiomodulation specifically. The better trials control for this, but not all do, so read the design before crediting the light.

A Reasonable Synthesis

Pulling the strands together, a fair-minded summary looks like this. There is a plausible biological mechanism. There is a body of mostly small, low-certainty trials. When those trials are sorted by dose, correctly dosed laser appears to produce clinically meaningful short-term pain relief, with the effect sometimes peaking weeks after treatment ends. The benefit is most consistent when laser is layered onto an exercise program. And the major guideline bodies have not yet incorporated any of this into a recommendation. All of those statements are true at once. The discomfort of holding them together is the actual state of the science.

Honest Evidence Grading

Stripping away the marketing, here is how the evidence stacks up by claim.

The umbrella reviews that grade these things using formal methods tend to land on low certainty. That is not the same as "it does not work." It means the trials are small, often poorly blinded, and inconsistent in how they report dose. When you see strong claims either for or against, be skeptical, because the underlying data does not support strong claims in either direction.

Where the Major Guidelines Stand

This is where reality bites for anyone hoping light is a settled treatment.

• The 2019 American College of Rheumatology / Arthritis Foundation guideline did not recommend laser therapy. It grouped laser with several modalities lacking sufficient evidence and offered no recommendation in favor (ACR/AF press release).
• The 2019 OARSI guideline for non-surgical management likewise did not endorse laser therapy as a recommended intervention; its strong recommendations centered on exercise, weight management, and self-management (OARSI 2019 guideline summary, APTA).

So the dose-aware positive meta-analysis exists, but the major bodies have not adopted laser into their recommendations. Part of that gap is timing, part is the low certainty of evidence, and part is that guideline panels weight large, well-blinded trials heavily and the laser literature is thin on those. If your physiotherapist offers laser, it is a defensible add-on, not a guideline-backed core treatment.

How It Compares to Other Options

It helps to put red light therapy in context against treatments that have more, or at least different, evidence.

For a broader view of light therapy and joint pain beyond the knee, see our reviews of red light therapy for pain relief and the clinical trials on red light therapy for joint pain and arthritis.

A few notes on the comparison table. Exercise and weight loss sit at the top not because they are glamorous but because they have the largest, most consistent evidence base and every guideline endorses them. Even modest weight loss reduces the load on the knee with every step, and strengthening the muscles around the joint takes pressure off the worn surfaces. No light device competes with that. Topical NSAIDs deserve a mention too: they relieve pain for many people with far less stomach and kidney risk than oral pills, which makes them a sensible first medication for a lot of patients.

High-intensity laser therapy is genuinely different from low-level laser despite the similar name. It uses far more power, generates heat, and is delivered by a trained operator. Network meta-analyses suggest it may outperform several other physical-therapy modalities, but it is a clinic-only treatment, not a home device, and its evidence base is still maturing. Do not conflate a HILT study with what a consumer red light panel can do.

Corticosteroid injections relieve pain reliably but only for weeks to a few months, and questions remain about repeated use over years. The point of the table is not to crown a winner. It is to show that red light therapy lives in the crowded middle, with real but modest and contested evidence, well below the well-supported basics and roughly alongside other physical-therapy adjuncts.

Safety Profile

This is the rare area where the news is clearly good. Across the trials, low-level laser and red light therapy show a strong safety record. Reported side effects are minimal and usually limited to mild, temporary warmth or redness at the treatment site. Serious adverse events are rare in the published literature.

The genuine cautions are practical:

• Eye protection. Lasers can damage the retina. Never look into a laser source; wear the goggles a clinic provides.
• Do not treat over cancerous lesions or suspicious skin growths without a doctor's sign-off, because of theoretical concerns about stimulating cell activity.
• Photosensitizing medications. If you take a drug that increases light sensitivity, check with your prescriber first.
• Pregnancy. Avoid treating directly over the abdomen; evidence in pregnancy is limited.

For a fuller rundown, our guide to red light therapy side effects goes deeper. The safety bar being low is part of why some clinicians offer it despite the modest evidence: the downside risk is small.

It is worth being clear about what FDA clearance does and does not mean here, because it gets misused in marketing. Most low-level laser and red light devices reach the U.S. market through the 510(k) clearance pathway as Class II devices, typically cleared for "temporary relief of minor muscle and joint pain, arthritis, and muscle spasm" and to increase local blood circulation. That clearance is about safety and being similar to devices already sold. It is not a finding that the device cures or even reliably treats osteoarthritis. A device can be perfectly legal to sell, genuinely safe, and still have only modest, contested evidence behind its pain claims. When you see "FDA cleared," read it as "safe and legally marketed," not "proven to work for your arthritis."

Who It Might Help, and Who Should Skip It

Reasonable candidate if you:

• Have mild to moderate knee osteoarthritis with ongoing pain despite exercise.
• Want a low-risk add-on to a physiotherapy program, not a replacement.
• Can access a clinic that uses a laser at a documented, WALT-aligned dose.
• Have realistic expectations: possible pain relief over weeks, not a cure.

Probably not worth it if you:

• Expect a home LED panel to replicate trial results (the trials used point-applied lasers).
• Have severe, bone-on-bone osteoarthritis where surgery is on the table.
• Are being told it regrows cartilage or reverses the disease. It does not.
• Cannot get any information on the device's wavelength or dose.

The single most useful question to ask a provider: what wavelength and how many joules per point? If they cannot answer, the 2019 data suggests you may be paying for the non-recommended-dose group, which performed no better than placebo.

How to Read a Red Light Study Without Getting Fooled

Because dose is the whole ballgame, a few habits help you separate signal from hype:

1. Check the wavelength and joules per point, not just "minutes of treatment." A study that omits dose is hard to trust.
2. Look for sham-controlled blinding. A fake-laser control group is the gold standard, since expectation alone relieves pain.
3. Watch the follow-up window. The 2019 analysis found relief often peaked two to four weeks after therapy ended, so a study that stops measuring on the last day may understate the effect.
4. Mind laser versus LED. Most positive trials used lasers; do not assume an LED panel performs identically.
5. Note whether exercise was included. Laser-plus-exercise studies tend to look better, and that benefit may be partly the exercise.

Frequently Asked Questions

Does red light therapy actually work for knee osteoarthritis?

It can, but only under specific conditions. The largest positive meta-analysis found meaningful pain relief, around 19 to 32 mm on a 100 mm scale, but only when the laser dose fell within recommended ranges. At the wrong dose, results matched placebo. Overall certainty of the evidence is rated low to moderate, and major guidelines do not yet recommend it.

Why do studies disagree so much?

Mostly because of dose. A 2015 meta-analysis pooled effective and ineffective doses together and found no benefit. A 2019 meta-analysis separated trials by dose and found a clear effect only in the correctly dosed group. The disagreement is less a true contradiction and more a lesson about how averaging hides the signal.

Can a home LED panel replace a clinic laser for my knee?

There is no good evidence for that. Nearly all knee osteoarthritis trials used point-applied lasers at a measured dose to multiple spots on the joint, not full-body LED panels. A home panel may still be safe and worth trying, but you cannot assume it reproduces the trial results.

Will red light therapy repair my cartilage?

No human clinical evidence supports cartilage repair or disease reversal. The realistic, evidence-based goal is reduced pain and possibly better function over a few weeks of treatment. Any claim that light regrows cartilage or cures osteoarthritis is not supported by the trials.

Is it safe to use on an arthritic knee?

The safety record is strong. Side effects in trials are minimal, usually mild warmth or redness. The main precautions are wearing eye protection with lasers, avoiding treatment over cancerous lesions, and checking with a doctor if you take photosensitizing drugs or are pregnant.

This article is for general information and is not medical advice. Talk with a qualified healthcare provider before starting any treatment for knee osteoarthritis.