Does red light therapy work for joint pain? Short-term relief only

Red light therapy for joint pain shows short-term relief in trials, and that is where the human evidence stops. A 2025 systematic review screened 59,873 records across five databases and found only eight randomised trials of light therapy for orthopaedic pain that met its standards; across knee osteoarthritis, chronic low back pain, plantar fasciitis, kneecap pain, shoulder impingement and tennis elbow, pain and physical function improved, but the trials used such different light settings that their results could not be pooled into a single number [15].

Where a trial put the two wavelengths head to head, the invisible one usually won. Photobiomodulation, the term researchers use for treating tissue with low-powered light, splits into visible red at roughly 630 to 660 nanometres on the scale that describes a light wave's length, and near-infrared at roughly 780 to 980 nanometres, which the eye cannot see but which travels deeper into tissue. In 30 patients with jaw muscle pain treated twice a week for five weeks, red light at 635 nanometres cut pain scores by about 62% and near-infrared at 980 nanometres by about 73%, with mouth opening widening 7.6 millimetres on red and 10.2 millimetres on near-infrared [4]. That trial had no placebo group.

The pattern repeated in the shoulder. Seventy-five people aged 45 to 70 with rotator cuff tendinopathy, meaning wear and irritation in the tendons that hold the shoulder in its socket, were split into five groups of 15 and given ultrasound, red light, near-infrared light, or either light combined with ultrasound, across 12 sessions in four weeks. Every group improved, and against ultrasound alone, red light came out ahead on 7 of 12 measures, near-infrared on 9, and near-infrared combined with ultrasound on all 12 [2].

The trials that used red light by itself are the least controlled. A wearable 660-nanometre neck device was tested on volunteers who all received the device, with nobody as a comparison: pain and sleep quality improved after two weeks, neck mobility after four, and no participant reported a side effect [1]. In 30 people with temporomandibular disorder, pain in the jaw joint and the muscles that move it, red light, near-infrared light and a near-infrared laser all reduced pain and widened mouth opening across eight sessions in four weeks, but no group beat any other and there was no dummy treatment to show the light itself was responsible [3].

One placebo-controlled test exists for sudden joint pain. Eighteen people were treated 8 to 12 hours after a hip replacement with a device combining red, near-infrared and laser diodes, nine getting the real thing and nine a dummy, and pain scores plus two of three inflammation markers in the blood fell in the treated group while the third did not shift [6].

Light at these wavelengths is absorbed inside cells rather than simply warming the skin, and in the laboratory what it changes is inflammation.

In human cartilage cells provoked with interleukin-1 beta, an inflammatory messenger the body releases in an irritated joint, near-infrared laser light lowered the production of three inflammatory proteins and quietened NF-kB, the switch that turns inflammatory genes on [10]. A different experiment on human cartilage cells found light raised the building blocks of cartilage, collagen type II and aggrecan, and unlike ultrasound it did not increase the enzymes that break cartilage down [8]. Those cells came from one donor.

The same effect shows up in whole animals. Near-infrared light aimed at the jaw joint of rats before an inflammatory injection cut the influx of immune cells, lowered three inflammatory messengers and raised an anti-inflammatory one [5]. In mice with surgically induced osteoarthritis, light slowed cartilage breakdown and thickening of the joint lining by pushing immune cells called macrophages out of their inflammatory state [7].

The longest animal test used 160 rabbits with surgically damaged knees. Six weeks of laser three times a week reduced how much the animals shifted weight off the sore leg and calmed inflammation in the joint lining, and eight weeks reduced visible cartilage damage across more of the joint surface [12].

Cartilage protection has been shown in cells, rabbits and mice, and never in a person [7][8][12]. The human trials measured pain, movement and questionnaires over weeks, so nothing in this evidence says a worn joint ends up structurally better.

The trials are also small and frequently uncontrolled. Thirty patients [3][4], eighteen patients [6] and one volunteer group with no comparison arm [1] carry most of the published red-light pain evidence, and a survey of 422 papers in this field concluded it still needs standardised protocols, dose-response work and long-term structural outcomes [9].

This evidence does not apply to:

• anyone expecting a worn joint to be rebuilt, because no human trial measured joint structure
• home panels and masks, because the trials used clinic lasers and LED arrays at specified energies per point
• pain whose cause has not been identified, because every trial recruited patients with a diagnosis
• children and teenagers, who were excluded from the systematic review's trials [15]

A review of non-drug treatments for long-running musculoskeletal pain puts low-level laser among the short-term options for neck and low back pain, and notes that exercise has the stronger support and that no single therapy works consistently [13].

Longevity clinics in the Longevinsights directory list red light therapy next to services like cryotherapy and IV drips, delivered as a panel or bed aimed at the skin with nothing injected. A first session means sitting or lying with the device pointed at the sore area for a few minutes per spot, and the trials give a sense of the commitment: 8 to 12 sessions across four or five weeks, two or three times weekly, was the standard schedule [2][3][4]. Nobody in the neck-pain pilot reported a side effect [1]. What the trials cannot say is whether relief outlasts the weeks they ran, or whether light aimed at a knee reaches the joint at all, given that the stronger results came from the deeper, invisible wavelength [4].