- The viral claim comes from two small studies at University College London: 24 people in the first, and a main test group of 20 in the second. They measured colour contrast (how faint a colour difference you can spot), not the sharpness most of us mean by "eyesight"
- The best result: one 3-minute dose of 670 nm deep red light in the morning improved blue-axis colour contrast by 17% on average three hours later. In 10 people retested a week later, the gain was 10%. An afternoon dose did nothing
- There was no sham light, and the authors said "caution is needed" in interpreting the data
- One eye use is FDA-authorised: the Valeda device for some people with dry macular degeneration (November 2024). The FDA called its benefit, about one line on the eye chart, of "marginal clinical significance"
- Device type matters. Red light slowed short-sightedness in children in a trial, but a laser device for myopia has been linked to retinal damage, and in a 2026 lab test two laser devices, including that one, reached a safety limit in under 3 seconds
The video goes like this: look at a red light for a few minutes each morning and your eyesight will get better. It's catnip for anyone over 40 holding the menu at arm's length, and it rides a boom in red-light panels and face masks sold for skin, muscles and "cellular energy".
Like many health trends, it starts from real research, then stretches it. Here's what the studies tested, where red light has earned a place in clinics, where it has caused harm, and what to do. It isn't medical advice: if your vision has changed, see an eye doctor.
Where the Viral Claim Comes From
Evidence: Preliminary — small studies without a sham lightThe idea comes from Glen Jeffery's lab at University College London (UCL). The reasoning: mitochondria, the parts of cells that make energy (ATP), decline with age, and earlier work in mice, bumblebees and fruit flies suggested long-wavelength light could partly restore that energy and the cells' function (Shinhmar, Sci Rep 2021; UCL, 2020).
The 2020 study: two weeks of daily light
In the first human test, 24 people aged 28 to 72 with no eye disease looked into a 670 nm deep red beam from a small LED torch for three minutes a day for two weeks. In people under about 40 it made no difference. In those around 40 and over, colour contrast sensitivity improved by "up to 20% in some people", mostly on the blue axis, and sensitivity in low light (rod function) also improved significantly (UCL; Shinhmar, J Gerontol A 2020). Improvements on the red axis weren't statistically significant.
Jeffery called the technology "simple and very safe", but UCL noted in 2020 that there were "currently no commercially available torches of the same specification".
The 2021 study: one morning dose
The study behind the "morning" part of the claim, published in Scientific Reports in November 2021, tested a single exposure (Shinhmar, Sci Rep 2021). Participants aged 34 to 70 looked for three minutes into a nine-LED device fitted over one eye with a tube. The light at the eye was about 8 mW/cm², a fifth of the roughly 40 mW/cm² used in earlier studies.
- Morning (8 to 9 AM), tested 3 hours later, 20 people: colour contrast thresholds improved by 17% on the blue (tritan) axis and 12% on the red (protan) axis on average.
- Morning dose, retested a week later, 10 people: 10% (blue) and 8% (red). Some people had returned to baseline; others hadn't.
- Afternoon (12 to 1 PM), 6 people: no significant change.
- No light at all, 10 people: no change on retesting.
Colour contrast gain after one 3-minute dose, by timing
Shinhmar et al., Scientific Reports 2021. Bars show the average reduction (improvement) in colour contrast thresholds, measured on the blue-yellow (tritan) and red-green (protan) axes. Groups were separate and small; there was no sham-light group and no reported blinding. Visual acuity was not measured.
What "better vision" meant here
Colour contrast sensitivity is the smallest difference in colour a person can detect, a measure of the cone cells behind colour vision. It isn't visual acuity, the sharpness of letters on an eye chart, and neither paper reports acuity. The studies say nothing about needing reading glasses less.
Why it's too early to call it a treatment
- No placebo. Nobody looked into a dummy light, and the paper doesn't describe blinding. Tests like these rely on what a person reports seeing, so expectations could play a part.
- Tiny groups. The afternoon result came from six people; the one-week result from ten.
- The authors are cautious. They wrote that "some of the data are noisy" and "caution is needed in interpretation of our data". UCL said that improvement "can vary markedly between those of similar ages" and that the work "would require a larger sample size" (UCL, 2021).
- Built in-house. A co-author designed and built the devices. The authors declared no competing interests.
None of that means the finding is wrong. It's an early signal that hasn't been tested the way a treatment needs to be.
The One FDA-Authorised Eye Use
Evidence: Moderate — one sham-controlled trial, modest effectOn 4 November 2024 the FDA granted LumiThera's De Novo request for the Valeda Light Delivery System, classed as a light-based device for dry age-related macular degeneration (AMD) (FDA De Novo database). It is a clinic instrument that shines LED light at 590 nm (amber), 660 nm (red) and 850 nm (near-infrared) into the eye, for people with dry AMD whose best-corrected vision is between 20/32 and 20/70 and who don't have wet AMD. A treatment series is nine sessions of 250 seconds over three to five weeks; in the pivotal trial, patients repeated a series every four months for two years (FDA decision summary).
The evidence came from the LIGHTSITE III trial: double-masked and sham-controlled, with 100 patients (65 got light, 35 got sham). At 13 months, treated eyes read 2.6 more letters than sham, which wasn't statistically significant. At 21 months the gap was 3.8 letters, which was. On an eye chart, five letters is one line. The FDA described the result as "a statistically significant improvement which is of marginal clinical significance", and states that after about two years treatment provides "approximately one line of visual acuity" more than no treatment.
The FDA also noted the treated group had a higher rate of progression to wet AMD than the sham group, for unclear reasons, and that the benefit may not last after treatment stops. It still concluded that "the probable benefits outweigh the probable risks." The device isn't for people sensitive to yellow, red or near-infrared light or with a history of light-triggered conditions such as epilepsy or migraine, and its safety and effectiveness haven't been established in people under 50.
The lesson: a modest benefit took a dedicated device, a strict schedule, a sham-controlled trial and two years to show, in one diagnosed disease, not in healthy eyes.
Red Light for Myopia, and the Safety Warnings
Evidence: Moderate for slowing myopia — safety unsettledThe other big use of red light on eyes is in children with short-sightedness (myopia). It's called repeated low-level red-light (RLRL) therapy. In a randomised trial published in Ophthalmology in 2022, 264 children aged 8 to 13 got either glasses alone or glasses plus a desktop device emitting 650 nm red light at 0.29 mW for a 4-mm pupil. They used it for 3 minutes, twice a day at least 4 hours apart, 5 days a week (Jiang, Ophthalmology 2022).
After a year, the eyeball had lengthened by 0.13 mm on average with red light against 0.38 mm with glasses alone, and refractive error had worsened by 0.20 versus 0.79 dioptres. There were no severe adverse events. But the control group got no placebo light, and the American Academy of Ophthalmology's summary noted it's unknown whether the results apply to other devices, wavelengths or powers (AAO).
Then came the safety reports
In 2023, JAMA Ophthalmology published the case of a 12-year-old girl who developed vision loss in both eyes after five months of RLRL therapy with an Eyerising device. Her vision fell from 20/20 to 20/30, and scans showed damage to the light-sensing layer at the centre of the retina. Three months after stopping, the damage had only partly recovered, with vision at 20/25 (Liu, JAMA Ophthalmol 2023, as reported by Myopia Profile). One case can't show how often this happens, only that it can.
In a study published in February 2026, University of Houston researchers Lisa Ostrin and Alexander Schill tested four myopia devices in the lab: three lasers and one LED. Against a recommended 180-second session, the Eyerising laser reached the ANSI group 1 safety limit (7-mm pupil) in 1.4 seconds and the Sky-n1201 laser in 2.8 seconds. A third laser, Future Vision, took 253 seconds or longer. The LED device, AirDoc, took 22,761 seconds (Optometry Times; Review of Optometry). Ostrin warned that "rapid clinical adoption has outpaced thorough safety validation", and China has recently reclassified red laser devices as Class III (University of Houston).
The clearest lesson in this field: "red light" covers devices that reach a safety limit thousands of times faster than others. A low-power LED and a laser aimed into a child's pupil are not the same thing.
What's been tested vs what's claimed
| Claim or use | Who was studied | Result | Verdict |
|---|---|---|---|
| "Red light improves eyesight" | 20 adults 38–70, one 3-min 670 nm LED dose | Blue-axis colour contrast 17% better at 3 h; 10% at 1 week (10 people); afternoon dose: no change sharpness not measured | Preliminary no sham light |
| Daily red light for ageing eyes | 24 adults 28–72, 3 min a day for 2 weeks | Colour contrast up to 20% better in some over-40s; low-light sensitivity improved; no effect under 40 | Preliminary small, no sham reported |
| Dry macular degeneration | 100 patients, sham-controlled (Valeda) | 3.8 letters better than sham at 21 months; 2.6 at 13 months not significant at 13 months | FDA-authorised "marginal clinical significance" |
| Myopia in children | 264 children 8–13, 650 nm device, 3 min twice a day | Eye lengthened 0.13 mm vs 0.38 mm in a year | Works in trials safety questions with lasers |
| Acne (home LED devices) | 6 trials, 216 people | Fewer inflamed and non-inflamed spots than controls; minimal side effects | Helps small trials |
| Muscle performance | 39 trials, 861 healthy people | Some benefit for fatigue and performance measures | Weak very low to moderate quality |
| Mouth sores in cancer care | MASCC/ISOO guideline review | Recommended to prevent oral mucositis in stem cell transplant and head and neck radiotherapy | Guideline-backed specific clinic protocols |
Shinhmar, Sci Rep 2021 and J Gerontol A 2020 (via UCL); FDA decision summary DEN230083; Jiang, Ophthalmology 2022; Ershadi, JAMA Dermatol 2025 (via Mass General Brigham); Vanin, Lasers Med Sci 2018; MASCC/ISOO, Support Care Cancer 2019.
Panels, Masks and the Rest of the Hype
Evidence: Mixed — best for specific clinical usesThe scientific name for red and near-infrared light therapy is photobiomodulation. It has a few areas of solid evidence and many claims that run ahead of the trials.
Cancer care. The strongest endorsement is unexpected. The Multinational Association of Supportive Care in Cancer and the International Society of Oral Oncology (MASCC/ISOO) recommend photobiomodulation to prevent oral mucositis, the painful mouth sores caused by cancer treatment, in people having stem cell transplants or head and neck radiotherapy (with or without chemotherapy). Clinicians are told to follow every parameter of a proven protocol, and for treating sores once they appear, the evidence wasn't strong enough for a guideline (MASCC/ISOO, Support Care Cancer 2019).
Skin. For home devices, the best evidence is in acne. A 2025 meta-analysis in JAMA Dermatology pooled 6 randomised trials with 216 people aged 12 to 50 with mild-to-moderate acne. Red and/or blue LED devices reduced inflamed and non-inflamed spots more than controls, with minimal side effects such as mild dryness. The authors called for larger trials (Mass General Brigham). Claims about wrinkles and collagen are a separate question, and not one this article sets out to settle.
Muscles. A 2018 review of 39 trials with 861 healthy people found "some effect in favor" of light therapy for measures such as time to exhaustion, but rated the evidence "very low to moderate" quality, citing small samples and varied protocols (Vanin, Lasers Med Sci 2018).
The pattern matches the eye research: results come from a specific wavelength, dose and schedule, and a panel or mask bought online may not match any of them.
What You Can Reasonably Do
For healthy eyes, red light is unproven. If you're still curious:
- Start with an eye exam, not a gadget. The American Academy of Ophthalmology recommends that adults with no symptoms or risk factors get a baseline eye disease screening at 40 (AAO). A change in your vision deserves a diagnosis.
- Know what the studies used: a low-power 670 nm LED at about 8 mW/cm² at the eye, for three minutes, in the morning. A product labelled "red light" may match none of that.
- Never look into a laser, including red-light devices sold for myopia. The two devices that hit safety limits in under 3 seconds were lasers.
- Panels and masks are made for skin. Don't stare into them. If a device comes with eye protection, wear it.
- Keep children off red-light eye devices unless an eye-care professional is supervising treatment and monitoring the retina.
- If you have an eye condition, such as macular degeneration, diabetic eye disease or glaucoma, talk to your ophthalmologist before trying any light device.
- Check your medicines. The FDA-authorised device isn't used within 30 days of drugs that make you sensitive to its wavelengths. Ask your pharmacist if you take anything photosensitising.
If what you want is better light for your body clock, daylight is free: our morning sunlight guide covers what it does for sleep.
Frequently Asked Questions
An interesting signal, not an eyesight fix.
The research behind the viral claim is real but small. Low-power deep red light improved colour contrast in adults over about 40 for hours to a week, in studies with no sham light. Nobody has shown it sharpens eyesight. The only FDA-authorised eye use is a clinic treatment for dry macular degeneration, with a modest benefit.
The risk is in the details: some laser devices exceed safety limits in seconds. Get your eyes examined, never look into a laser, keep kids away from these devices outside supervised care, and talk to your eye doctor first.
Sources
- Shinhmar H, Hogg C, Neveu M, Jeffery G. Weeklong improved colour contrasts sensitivity after single 670 nm exposures associated with enhanced mitochondrial function. Scientific Reports 2021;11:22872, published 24 November 2021 (doi:10.1038/s41598-021-02311-1).
- University College London. Morning exposure to deep red light improves declining eyesight. 24 November 2021.
- Shinhmar H, Grewal M, Sivaprasad S, et al. Optically improved mitochondrial function redeems aged human visual decline. Journals of Gerontology: Series A 2020;75(9):e49–e52 (doi:10.1093/gerona/glaa155).
- University College London. Declining eyesight improved by looking at deep red light. 29 June 2020.
- US Food and Drug Administration. De Novo database entry DEN230083: Valeda Light Delivery System (decision date 4 November 2024) and De Novo decision summary.
- LumiThera. LumiThera Obtains FDA Authorization of Valeda Treatment for Dry AMD Patients to Improve Vision. Business Wire, November 2024.
- Jiang Y, Zhu Z, Tan X, et al. Effect of repeated low-level red-light therapy for myopia control in children: a multicenter randomized controlled trial. Ophthalmology 2022;129(5):509–519 (doi:10.1016/j.ophtha.2021.11.023). Caveats as summarised by the American Academy of Ophthalmology (28 January 2022).
- Liu H, Yang Y, Guo J, Peng J, Zhao P. Retinal damage after repeated low-level red-light laser exposure. JAMA Ophthalmology, published online 25 May 2023. As reported by Myopia Profile.
- Ostrin LA, Schill AW. Safety evaluation of 4 red light therapy devices for myopia. JAMA Ophthalmology 2026;144(3):255–258, published online 5 February 2026 (doi:10.1001/jamaophthalmol.2025.5660). As reported by Optometry Times (9 February 2026), Review of Optometry and the University of Houston (2 March 2026).
- Zadik Y, Arany PR, Fregnani ER, Bossi P, et al.; Mucositis Study Group of MASCC/ISOO. Systematic review of photobiomodulation for the management of oral mucositis in cancer patients and clinical practice guidelines. Supportive Care in Cancer 2019;27(10):3969–3983 (doi:10.1007/s00520-019-04890-2).
- Ershadi S, Barbieri JS. At-Home LED Devices for the Treatment of Acne Vulgaris: A Systematic Review and Meta-Analysis. JAMA Dermatology 2025, published online 5 March 2025 (doi:10.1001/jamadermatol.2025.0019). As reported by Mass General Brigham (March 2025).
- Vanin AA, Verhagen E, Barboza SD, Costa LOP, Leal-Junior ECP. Photobiomodulation therapy for the improvement of muscular performance and reduction of muscular fatigue associated with exercise in healthy people: a systematic review and meta-analysis. Lasers in Medical Science 2018;33(1):181–214 (doi:10.1007/s10103-017-2368-6).
- American Academy of Ophthalmology. Get an Eye Disease Screening at 40.