Photobiomodulation

Mester's observation in 1967

A 2016 commentary by Michael Hamblin dates the discovery of what became low-level laser therapy to 1967, when Endre Mester at Semmelweis Medical University in Hungary tried to repeat an experiment by Paul McGuff of Boston, who had used the newly invented ruby laser against tumors in rats. According to the commentary, Mester's laser was underpowered and did not destroy the tumors, but the treated rats showed a higher rate of hair growth and better wound healing.

The same commentary says later work found that non-coherent light-emitting diodes (LEDs) with comparable parameters performed equally well as low-power lasers in the studies it describes, so a laser was not required. This is a short historical summary by one author who works in the field, and it does not reconstruct the clinical history that followed.

A field with many names

A 2015 editorial by Juanita Anders, Raymond Lanzafame and Praveen Arany lists earlier labels including laser biostimulation, low-level laser therapy, low-intensity laser therapy, cold laser and photobiomodulation. It reports that a nomenclature consensus meeting organized by the North American Association for Light Therapy and the World Association for Laser Therapy in September 2014 concluded that photobiomodulation better reflects the scientific basis of the therapy.

The U.S. National Library of Medicine's MeSH vocabulary records the descriptor Low-Level Light Therapy as introduced in 2002. Before that, material of this kind was indexed under therapeutic uses of lasers from 1983 to 2001.

Spread to consumer devices

A 2026 study of social media posts reports that the popularity of red light therapy has grown substantially and that social media has played a significant role. The sources reviewed for this page do not trace when clinic-based laser use gave way to home LED devices, and they do not give dated milestones for that shift.

From low-level laser therapy to photobiomodulation

Hamblin gives three reasons for the change of name: lasers were shown not to be necessary, 'low-level' was considered a subjective term, and both inhibition and stimulation of biological processes could be therapeutically useful. Anders and colleagues make a similar argument, calling 'low' and 'level' vague and 'laser' obsolete now that LEDs and broadband light are used.

The 2015 editorial proposes a definition of photobiomodulation therapy as a form of light therapy using non-ionizing light sources, including lasers, LEDs and broadband light, in the visible and infrared spectrum, producing non-thermal effects. It reports that photobiomodulation therapy was added to MeSH as an entry term for the 2016 version, with indexing from November 2015. The current MeSH record lists photobiomodulation among the entry terms for Low-Level Light Therapy, with a scope note that the effects are a response to the light and not due to heat.

Red light therapy and related labels

MeSH entry terms include LLLT, laser biostimulation, laser phototherapy and low-power laser therapy. 'Red light therapy' is the label most often used for consumer products. In a 2026 analysis of 132 social media posts, most recommended devices combined red and near-infrared light, so the label does not reliably indicate a single wavelength.

In the research literature, photobiomodulation can also cover wavelengths outside red and near-infrared. Hamblin describes blue and green light as showing potential but facing significant limits in how far they penetrate tissue. Studies of one wavelength, dose or device should not be assumed to apply to the others.

Light sources and wavelengths

Hamblin's 2017 review describes the effective wavelengths as red light at 600 to 700 nm and near-infrared light at 770 to 1200 nm, with optimal tissue penetration at about 810 nm. He describes the range between 700 and 770 nm as likely to give disappointing results. A position paper from the World Association for Photobiomodulation Therapy (WALT) gives wavelengths in the supportive cancer care literature as roughly 600 to 1000 nm.

Dose, devices and sessions

The WALT paper summarizes reported power density as 5 to 150 mW/cm2 and energy density as 0.1 to 12 J/cm2, with application of about 30 to 60 seconds per point. For oral mucositis it gives preventive intraoral protocols using visible wavelengths of 630 to 680 nm, and treatment protocols repeated three to four times weekly for 15 to 20 sessions. It states that combining parameters from different protocols changes outcomes unpredictably and that a one-size-fits-all or point-and-shoot approach should no longer be acceptable.

The same paper notes that a device's real output often differs from the manufacturer's specification, so power should be measured routinely. In knee osteoarthritis, a meta-analysis expressed dose per treatment spot, reporting benefit at 4 to 8 J for 785 to 860 nm light and at 1 to 3 J for 904 nm light. These figures come from specific trials and are not general settings.

Uses studied in clinical settings

Hamblin describes photobiomodulation as the use of red and near-infrared light to stimulate healing, relieve pain and reduce inflammation. The WALT position paper lists complications of cancer treatment for which it outlines photobiomodulation protocols, including oral mucositis, radiodermatitis, dysphagia, xerostomia, taste changes, trismus, lymphedema, hand-foot syndrome, alopecia, chronic graft-versus-host disease and peripheral neuropathy.

Other uses with systematic reviews include knee osteoarthritis pain and androgenetic alopecia, covered below. A 2026 narrative review also discusses red and near-infrared light for dry age-related macular degeneration, dry eye and myopia, and concludes that for most ophthalmic indications the clinical evidence is heterogeneous and insufficient to define photobiomodulation as established therapy.

Consumer and social media claims

In the 2026 analysis of 132 social media posts, skin was the benefit named most consistently, at 88.6 percent of posts. Physician posts focused on anti-aging and acne, while non-physician posts named a broader range of targets. Only 8.3 percent of posts cited peer-reviewed articles. The authors conclude that claimed benefits of promoted at-home devices are often well beyond substantiated evidence.

No systematic review of photobiomodulation alone for skin aging was verified for this page. Two systematic reviews surfaced in searches, but they concerned LED-activated photodynamic therapy with a photosensitizing drug, a different intervention, and are not used here.

Oral mucositis in cancer care

The strongest guideline-level material concerns prevention of oral mucositis. A systematic review by the mucositis group of MASCC/ISOO (Multinational Association of Supportive Care in Cancer and International Society for Oral Oncology) made recommendations for preventing oral mucositis and related pain with photobiomodulation in hematopoietic stem cell transplantation and in head and neck radiotherapy, with or without chemotherapy. It reached no guideline for treating established mucositis or for chemotherapy-related mucositis because evidence was inadequate and clinical settings varied widely.

A 2026 meta-analysis of 30 randomized trials (1,748 patients with head and neck cancer) found fewer cases of severe oral mucositis (risk ratio 0.46, 95% CI 0.29 to 0.71) and severe oral pain (0.35, 0.23 to 0.53) than controls, with no significant effect on xerostomia or quality of life. The authors rate certainty as very low to low, with most studies at high risk of bias. This evidence concerns specific cancer treatment settings and should not be carried over to wellness uses.

Knee osteoarthritis pain

A 2019 meta-analysis of 22 placebo-controlled trials (1,063 participants) reported that pain fell by 14.23 mm on a 100 mm scale more than with placebo at the end of therapy, and by 18.71 mm in trials using WALT-recommended doses. No adverse events were reported. The authors noted that low-level laser therapy was not recommended in major knee osteoarthritis guidelines.

A 2024 meta-analysis of ten placebo-controlled trials (542 participants) rated all studies as unclear to high risk of bias. It found a reduction in pain at rest of 0.7 points (95% CI 1.1 to 0.2) but no significant effect on the Timed Up and Go test, with very low certainty. The authors state that this certainty does not allow them to recommend isolated use.

Hair growth

A 2024 meta-analysis of 38 studies (3,098 patients, mostly androgenetic alopecia) found greater gain in hair density than placebo after 4 to 26 weeks of low-level laser or LED therapy, with a standardized mean difference of 1.14 below 20 weeks and 1.44 above 20 weeks. Heterogeneity was high (I2 about 81 to 88 percent). Data were insufficient to pool results for other alopecia types.

A 2025 meta-analysis of four randomized trials comparing minoxidil plus low-level laser therapy with minoxidil alone found no significant difference in hair count (mean difference -0.04, 95% CI -1.22 to 1.14) or hair diameter. Whether the light adds benefit beyond a standard topical treatment is therefore not established.

Cytochrome c oxidase and other chromophores

Hamblin identifies cytochrome c oxidase in the mitochondrial respiratory chain as the main identified chromophore, a molecule that absorbs the light. The prevailing theory is that light dissociates inhibitory nitric oxide from the enzyme, allowing increased respiration and ATP production. Opsins, light-gated ion channels and flavoproteins are proposed secondary targets, and a role for structured water is described as still hypothetical. These mechanisms come from laboratory and animal work and do not by themselves show clinical benefit in people.

The biphasic dose response

Hamblin describes a biphasic dose response in which a response peaks at some dose and then diminishes or disappears when the dose rises further. His example is that ATP and mitochondrial membrane potential in cortical neurons peaked at 3 J/cm2 of 810 nm laser light, with inhibitory effects at 30 J/cm2. The WALT paper reports a similar emphasis on dosimetry, and a knee osteoarthritis meta-analysis found benefit only within specific dose bands. In practice, results from a given wavelength and dose do not transfer to other settings.

Reported adverse effects

Hamblin describes an almost complete lack of reported adverse effects, provided the parameters are understood at least at a basic level. The WALT paper identifies one study that reported a burning sensation in 50 percent of pediatric patients, and it recommends that treatments remain non-thermal, below 45 degrees Celsius. It also notes that very high power and dose combinations could induce adverse effects in tumor models. Sparse adverse-event reporting in trials is not the same as demonstrated absence of harm.

Tumor sites, eyes and device quality

The WALT paper advises avoiding direct application over tumors and states that robust evidence about whether photobiomodulation protects or enhances malignant cells has not been published. People in cancer treatment who are offered it as supportive care should discuss it with their oncology team.

A 2026 narrative review reports generally favorable short-term tolerability in ophthalmic studies but says long-term safety remains insufficiently defined. The sources reviewed here give no eye-protection requirements for any device, so that question is left open. The social media analysis found that the output of promoted at-home devices is untested against the specifications in cited studies, which affects whether a consumer device delivers a dose resembling that in trials.

Photobiomodulation should not delay or replace necessary medical care, such as cancer treatment or evaluation of persistent hair loss or joint pain. Results for one condition, wavelength or dose cannot be assumed to apply to another, and the reviews summarized above do not support treating the light as a substitute for standard treatment.