Red light therapy: what the evidence covers, and what it does not

The mechanism, briefly

Red and near-infrared light, roughly 630 to 850 nanometres, penetrates tissue better than visible light at shorter wavelengths and is absorbed by cytochrome c oxidase in the mitochondria. The proposed consequence is a shift in ATP production and in nitric oxide and reactive oxygen species signalling. This is a real field with a real name — photobiomodulation — and it is not fringe. What it is not is a single effect with a single dose.

Where the evidence is strongest

Skin. Randomised trials support modest improvements in wrinkles, collagen density and wound healing, and several devices carry regulatory clearance for dermatological use. There is also reasonable support for some musculoskeletal pain. If you bought a mask for your face, you bought the best-evidenced version of the product, and the honest expectation is a modest change over months rather than a visible one over weeks.

Where it thins out

Muscle recovery and performance. The literature is large but built from small studies with wildly heterogeneous protocols, and the results do not converge. Dose is the core of the problem: wavelength, irradiance, duration and — critically — distance all vary between studies and between products, and irradiance falls off sharply with distance. “Twenty minutes at thirty centimetres” from one panel is simply not the same dose as the same instruction from another, which means the studies are not really testing the same intervention, and neither are you.

The timing problem nobody mentions

A red light panel is, before anything else, a bright light. Red is genuinely the least disruptive part of the spectrum for the body clock — the retinal cells that set circadian timing peak in sensitivity around 480 nanometres, far from red — so a red panel suppresses melatonin much less than a white ceiling light of similar brightness. But much less is not none, and a high-irradiance panel used at close range delivers a substantial light dose at exactly the hour when the point of the wind-down is to reduce light. If you use one in the evening, do not look into it, and count it as part of your light environment rather than treating it as exempt from the rules you are applying to your phone.

Testing it, and what to expect

Match the evidence to the method. Skin claims are answered by photographs in consistent lighting over months, not by a wearable. Recovery and sleep claims can be tested against your own history, but only in blocks: two weeks on, two weeks off, watching sleep latency and resting heart rate rather than any single night, and tagging the block so the comparison is legible later. Expect a small effect, because that is what the better studies report. On safety, do not stare into a panel — some are bright enough that the manufacturer supplies goggles, and that instruction exists for a reason. This is general information, not medical advice.

Frequently asked questions