Red light therapy
and eye health
In November 2024, photobiomodulation became the first non-invasive treatment ever authorised by the FDA for vision loss in dry AMD. Here is what the clinical evidence shows, where it is still unsettled, and what home panel users actually need to know.
A clinical milestone most people missed
Most conversations about red light therapy focus on skin, hair, and pain. But in November 2024, something more significant happened in a part of medicine that rarely gets consumer attention: photobiomodulation received FDA De Novo authorisation as the first ever non-invasive treatment for vision loss in dry age-related macular degeneration. No injections. No surgery. A targeted light device, applied to the retina, improving visual acuity in a clinical trial at 10 US retinal centres.
This is not a future possibility. The LumiThera Valeda system is now available at ophthalmology clinics, it has a billing code (CPT Category III, effective January 2025), and its extension trial data - presented at the Association for Research in Vision and Ophthalmology in May 2025 - shows benefits persisting out to 4.5 years. Alcon, one of the world's largest eye care companies, has since acquired LumiThera.
At the same time, a separate and more contested application - red light for myopia control in children - has accumulated impressive efficacy data alongside unresolved safety questions. And home panel users have their own separate question: what does any of this mean for how they use their device?
This blog works through all three areas with the same standard we apply across the NovaThera blog: accurate sourcing, honest limits, and no overclaiming.
Why the eye responds to red and near-infrared light
The retina converts light into neural signals and has one of the highest densities of mitochondria - the energy-producing structures inside cells - of any tissue in the body. This is why it responds to the same photobiomodulation mechanism that operates across every other application of red light therapy, but often more acutely, particularly when retinal cells are ageing or diseased.
How red light acts on retinal tissue - in plain terms
Light reaches the cells
Red and near-infrared wavelengths penetrate the eye and are absorbed by structures inside the mitochondria - the energy-producing parts of your cells. The retina has an unusually high concentration of mitochondria, which is why it responds to this type of light.
Energy production is restored
In ageing or diseased retinal cells, a molecule called nitric oxide builds up and blocks the cell's energy engine. Red light helps clear this blockage, allowing the cell to produce energy normally again. Cells that are already healthy are largely unaffected.
The retina benefits
Better cellular energy supports the survival of photoreceptors (the cells that let you see), reduces the build-up of fatty deposits that damage the macula in AMD, and lowers the inflammation associated with retinal disease.
Dry AMD is the leading cause of central vision loss in people over 55 in developed countries. Until 2024, there was no non-invasive treatment for it. Wet AMD could be treated with eye injections (anti-VEGF therapy), but dry AMD - which accounts for approximately 90% of all AMD cases - had no equivalent option beyond AREDS2 nutritional supplements.
The LIGHTSITE III trial changed that picture. Conducted at 10 leading US retinal centres and enrolling 100 patients with early to intermediate dry AMD, it was a prospective, double-masked, randomised controlled trial comparing the LumiThera Valeda system against sham treatment. The Valeda delivers three wavelengths simultaneously: yellow (590 nm, 5 mW/cm²), red (660 nm, 65 mW/cm²), and near-infrared (850 nm, 8 mW/cm²), administered as nine sessions over 3-5 weeks, repeated every four months.
The three figures below are from the published trial results. Each one tells a different part of the story.
The primary endpoint - best-corrected visual acuity gain - was met, with a statistically significant difference between the PBM and sham groups (between-group difference 2.4 letters, p=0.02). At 13 months, PBM-treated eyes showed a mean gain of 5.4 letters, compared to 3.0 letters in the sham group. 55% of PBM-treated eyes gained five or more letters. The effect on disease progression was more striking: new geographic atrophy occurred in 9.8% of sham-treated eyes but only 1.1% of PBM-treated eyes (OR 9.4, p=0.024).
The extension trial (LIGHTSITE IIIB), presented at the Association for Research in Vision and Ophthalmology in May 2025, re-enrolled 36 participants (63 eyes) from LIGHTSITE III for four additional treatment cycles over 13 months, following a 20-month treatment gap. Patients who had originally received PBM regained approximately 5 letters after retreatment began. Patients from the original sham arm who were switched to PBM stabilised but did not show the same recovery, suggesting earlier treatment produces better long-term outcomes. Analysis across both studies shows PBM-treated eyes sustaining cumulative benefit over approximately 4.5 years.
The FDA authorisation was granted under the De Novo pathway, establishing Valeda as the first device in a new class of ophthalmic PBM devices. An important nuance: this is an authorisation, not a full approval. Insurance coverage is still developing, and some providers may classify the treatment as investigational. But the clinical evidence - across three LIGHTSITE trials and now an extension study - is the most rigorously designed PBM dataset in any ocular application.
What this means in practice
If you or a family member have been diagnosed with early to intermediate dry AMD, PBM treatment is now a clinical option worth discussing with your ophthalmologist. The Valeda system is CE-marked in the UK, UKCA-marked, and available at specialist retinal centres. It is not something you replicate at home - the device uses calibrated multiwavelength delivery at specific irradiance levels that consumer panels do not replicate. But the underlying biology is genuine, and the trial evidence is the strongest in the history of PBM in ophthalmology.
Repeated low-level red light therapy (RLRL) for myopia control in children is the other significant eye health application in the research. The efficacy data is striking. A 2022 multicentre randomised trial published in Ophthalmology - the flagship journal of the American Academy of Ophthalmology - enrolled 264 children aged 8-13 with myopia and found that RLRL treatment reduced axial elongation from 0.38mm to 0.13mm over 12 months, compared to single-vision spectacles alone. A subsequent meta-analysis in Scientific Reports (2025) covering multiple RCTs concluded that RLRL outperforms both orthokeratology lenses and 0.01% atropine for axial length control.
The mechanism differs from the AMD application. Rather than targeting diseased retinal cells, RLRL is thought to act on choroidal thickness - the vascular layer behind the retina. Increased choroidal thickness appears to slow the axial elongation that drives myopia progression.
The two cards below summarise where the research currently stands - one on how well it works, one on what safety questions remain open.
axial elongation (eye lengthening that drives myopia) in the RLRL group over 12 months, vs 0.38mm in children wearing standard spectacles alone - a 66% reduction in myopia progression rate
commercially available RLRL devices tested hit internationally recognised retinal safety thresholds before the 3-minute treatment was even complete - meaning some devices may be delivering more light energy to the retina than current safety guidelines recommend
The safety picture is where this application becomes actively contested. A 2025 study in JAMA Ophthalmology by Liao et al. examined 52 children who had used RLRL therapy for at least one year, comparing them to 47 controls. Using adaptive optics scanning laser ophthalmoscopy - high-resolution retinal imaging capable of resolving individual photoreceptor cells - they found significantly decreased cone density within 0.5mm of the fovea in the RLRL group (mean difference of 2.1 × 10³ cells/mm² at 0.3mm temporal eccentricity, p=0.003). Some children also showed drusen-like lesions. No functional vision loss was detected, and the authors acknowledged that longer-term follow-up was needed to determine whether the changes have clinical significance.
The device safety issue is separately documented. Ostrin and Schill (Ophthalmic Physiology and Optics, 2024) measured the actual optical output of four commercially available RLRL devices using calibrated radiometers. All four reached ANSI retinal irradiance limits within exposure times shorter than the recommended 180-second treatment time. This is significant because the devices are classified as Class I laser products - a classification based on accidental rather than intentional, repeated exposure. China has since reclassified RLRL laser devices from Class I to Class III. None of the devices used in these protocols are FDA-cleared for home use in any market.
The most balanced current summary comes from a 2024 systematic review of RLRL safety covering 2,380 participants aged 3-18: no permanent vision loss was recorded; the most common side effect was a temporary afterimage resolving within 6 minutes; side effect incidence was 0.088 per 100 patient-years. But the review concluded that further adequately powered studies of longer duration are needed, and that routine OCT monitoring during treatment is essential to detect retinal changes early.
The honest position on RLRL for myopia
The efficacy signal is real and clinically meaningful - myopia is a growing global public health problem and current evidence places RLRL among the most effective controls available. But the safety picture in 2026 is not fully resolved, particularly regarding long-term cone photoreceptor health. This is a treatment that warrants clinical supervision, regular OCT monitoring, and should not be replicated with general-purpose consumer panels. The devices used in the research are purpose-built laser systems operating under specific protocols. They are not equivalent to home RLT panels.
The eye safety concerns documented in the myopia literature relate specifically to laser-based RLRL devices - desktop units that emit a coherent, focused 650nm laser beam directed into the eye deliberately, twice daily, for 3-minute sessions. This is a fundamentally different category from the LED panels used for skin, pain, and recovery applications.
Consumer LED panels emit incoherent, diffuse light across a broad beam. The irradiance at distance falls off quickly - an LED panel at 30cm from the face delivers a fraction of the energy per cm² that a clinical laser device delivers at the retina. No reported cases of retinal injury from standard LED red light therapy panels appear in the peer-reviewed literature. The American Academy of Ophthalmology notes that many ophthalmologists are cautiously optimistic about PBM while emphasising the need for more data.
That said, there are reasonable precautions that apply to all users.
✓ Use eye protection near the face
When positioning a panel near your face, close your eyes or wear the protective eyewear supplied with your device. NIR wavelengths (810-850 nm) are invisible and do not trigger the blink reflex, meaning your eye's natural defences are not activated. Opaque goggles or simply closing your eyes removes the risk entirely.
✓ Do not point a panel at open eyes
Staring into a high-power LED panel at close range is uncomfortable and unnecessary. The therapeutic light reaches the tissues you are treating without direct ocular exposure. This applies to any bright light source, not just RLT panels.
! Pre-existing eye conditions
People with a history of retinal disease, photosensitising medications, severe dry eye, or light sensitivity should discuss LED therapy near the face with their ophthalmologist before use. The general population safety profile is favourable; those with existing ocular vulnerability have less established data.
! Do not replicate clinical protocols
Neither the AMD Valeda treatment nor the myopia RLRL protocol can be safely replicated with a home LED panel. The AMD treatment uses a calibrated multiwavelength clinical device. The myopia protocol uses a laser system with an unresolved safety profile. If either applies to you, the appropriate route is clinical consultation.
The key distinction
Clinical PBM for the eye uses precisely calibrated laser or LED systems, at specific wavelengths, doses, and distances, applied by or under the supervision of an eye care professional. Consumer LED panels are a different category of device. The research base for each is separate. The safety considerations for each are separate. Do not conflate them.
Where things stand
Red light therapy's relationship with eye health sits in three distinct places at once. For dry AMD, the evidence is now at the level of a regulatorily authorised clinical treatment - the strongest tier possible. For myopia control in children, the efficacy data is compelling but the safety picture needs more time and larger trials before the field reaches consensus. For home panel users, the relevant question is not whether ocular PBM works clinically - it clearly can - but whether your LED panel is doing anything to your eyes in the first place. The answer, when used sensibly and with eyes protected, is no meaningful risk.
The broader point is that the eye is not a body part to be excluded from the story of photobiomodulation. It is mitochondria-rich, metabolically demanding, and in the case of AMD, one of the most underserved areas in medicine. The 2024 FDA authorisation is a significant moment. The science behind it has been building for over a decade through the LIGHTSITE trial series. And with Alcon now behind the device, larger studies are likely coming.
Sources
NovaThera
Panels built on verified science.
Published irradiance data. Verified wavelengths. Honest descriptions of what the research supports. If you want to understand how red light therapy fits into your routine, start with the evidence.