Red Light Therapy Wavelength and Dose
Two numbers decide whether a red light device can do anything at all: the wavelength it emits and the dose it delivers. Everything else on the box is decoration.
Most products sold for pets publish neither properly, which is why this is the single most useful thing to understand before buying one.
Why These Particular Wavelengths
Light does not do this by warming tissue. It does it by being absorbed by a specific molecule, cytochrome c oxidase, the last enzyme in the chain that cells use to make energy. That molecule absorbs some wavelengths well and others barely at all, and the useful bands are set by its absorption spectrum rather than by anything visual.
The mapping work that established this found four peaks in the biological response, at roughly 620, 680, 760 and 820 nanometres. That is why devices cluster where they do instead of spreading evenly across the red end of the spectrum.
The working bands are about 600 to 700 nm and about 780 to 1100 nm. The gap between them is not an oversight. The range from 700 to 780 nm coincides with a trough in the enzyme's absorption and is considered largely ineffective.
At the same time, tissue has its own preferences. Blood absorbs strongly at the short end and water absorbs strongly at the long end, which leaves an optical window running from roughly 650 to 1350 nm where light travels furthest. The bands that are useful biologically sit either side of the range that transmits best physically, which is exactly why real devices carry two wavelengths rather than one.
Red and Near Infrared Do Different Jobs
A wavelength around 650 nm is visible red. It is absorbed near the surface and it is the one that matters for skin, coats, wounds, surgical sites and anything you can see.
A wavelength around 808 or 810 nm is near infrared and invisible. It scatters further before it is absorbed, which is why devices aimed at joints and soft tissue carry it. If a product offers only visible red light and claims to treat a hip, the physics is not on its side.
A device with both is covering two different depths, and the sensible reading of a two wavelength specification is that it is trying to.
How Deep It Actually Goes
Here the published figures disagree with each other by an order of magnitude, and any page that gives you one confident number is hiding that.
Measured in people, light at 660 nm spread a mean of 14 to 21 millimetres through tissue and light at 830 nm spread 20 to 26 millimetres, with skin pigmentation making a highly significant difference. Other work reports that a diode laser passed only about eleven percent through under two millimetres of skin.
Both can be true, because they are measuring different things. Some photons arriving at a depth is not the same as a therapeutic dose arriving at that depth, and the second is what matters. The same source calculates a device delivering roughly one hundred and fortieth of the minimum useful dose to its intended target.
The other figure worth carrying: hair alone can absorb around ninety eight percent of what a device emits before any of it reaches skin.
Coat Colour Changes the Answer
This is the dog specific finding that almost nothing written for owners mentions, and it is the most practically useful thing on this page.
In a study of 47 dogs across black, brown and white coats, coat colour significantly affected how much light was transmitted, with white transmitting most and black least. Clipping the area significantly improved transmission. Coat length, counter intuitively, was not a significant predictor.
So a black labrador and a white terrier treated identically are not receiving the same dose, and parting or clipping the coat matters more than treating for longer. If you are going to change one thing about how you use a device, change that.
What a Dose Is
Three quantities, and they are related by simple arithmetic.
| Quantity | Unit | What it describes |
|---|---|---|
| Optical output | mW | How much light leaves the device |
| Irradiance | mW/cm² | How concentrated it is on the skin |
| Fluence, or energy density | J/cm² | Irradiance multiplied by time, the dose itself |
One watt is one joule per second, so a device putting out 500 mW for ten minutes emits 300 joules in total. That number alone tells you nothing until you know what area it was spread over.
The published dose guidance is specific. The World Association for Photobiomodulation Therapy recommends 4 to 8 joules per treatment point for tendon problems and 4 to 16 for arthritis in the 780 to 860 nm band, with a maximum irradiance of 100 mW per square centimetre, and states that doses outside those windows should not be considered low level laser therapy at all.
Veterinary sources land in a similar place, generally 2 to 6 joules per square centimetre for pain, 8 to 10 for inflammation and 8 to 12 for most joints.
One caveat that matters for dogs: those dosage tables state they are based on white human skin. Nobody has published an equivalent for a pigmented, coated animal, so applying them to a black dog is an extrapolation rather than a protocol.
Peak Output Is Not Average Output
This is the specification trick that catches the most buyers, and it is entirely legal because both numbers are true.
A pulsed device produces its stated maximum only at the top of each pulse. For the rest of the cycle it produces much less, and the energy delivered over a session follows the average rather than the peak. A device quoting 500 mW peak can be delivering a small fraction of that on average, which means the joule arithmetic below applies only if the figure you started with was the average one.
So the question to ask a seller is not how many milliwatts, it is whether that figure is peak or average. A seller who cannot answer has told you something useful anyway.
Why Longer Is Not Better
The intuition that a bigger dose does more is wrong here, and it is wrong in a specific, documented way.
The response follows a curve that rises and then falls. Work reviewing the biphasic dose response in low level light therapy collected examples: cell proliferation peaking at 0.88 joules per square centimetre and reduced at 8.68, wound healing best at 2 and worse at 50, and a model in which raising irradiance tenfold while delivering the same total energy eliminated the benefit completely.
Read that last one again, because it is the important one. The same number of joules delivered faster gave a different and worse result. Total energy is not the variable on its own.
At the high end the review describes beneficial signalling molecules falling while potentially harmful ones rise. There is a real ceiling, and running double sessions because a normal one did nothing is the most common way owners walk through it.
Laser or LED
A long running argument, and the weight of evidence now sits on one side of it.
Comparative work matching wavelength and dose has generally found LEDs and laser diodes producing similar results, which implies coherence is not the active ingredient. A review of the question concludes that LED photobiomodulation is here to stay, citing the absence of laser safety requirements, the ability to treat a large area at once and a much lower cost per milliwatt.
A minority position holds that only lasers qualify. It exists, it is cited, and it is not where the field has landed.
The practical consequence is that the light source type is a weaker signal than the wavelength and the dose, and a device that publishes the first but not the other two is telling you the least useful thing about itself.
The Spec Sheet You Should Demand
- Wavelengths in nanometres, as numbers, not the words red and infrared
- Total optical output in milliwatts at the aperture
- Treatment area in square centimetres, or irradiance at a stated distance
- Laser class
- Session timer range and any adjustable power levels
The fourth and fifth items are safety. The first three are what let you calculate a dose, and without all three you cannot.
The Arithmetic, Worked Through
Take a device emitting 500 mW, run for ten minutes. That is 300 joules of total energy. Now spread it:
| Area treated | Dose delivered | Where that sits |
|---|---|---|
| 10 cm² | 30 J/cm² | Past the point where the response starts falling away |
| 50 cm² | 6 J/cm² | Within the usual veterinary range |
| 100 cm² | 3 J/cm² | Low end, analgesic range |
Same device, same ten minutes, a tenfold swing in delivered dose purely from how much of the dog you covered. That is why moving the head steadily over an area matters more than holding it on a spot, and why treatment area belongs on the specification.
It also explains why the clinic and home versions of this treatment are harder to compare than the power figures suggest, which our guide to clinic laser therapy and what a course costs goes into.
What the Market Actually Publishes
There is no published audit of how many consumer pet devices disclose their specifications, so anyone quoting you a percentage has made it up, including us.
What has been measured is worse than a disclosure problem. A pilot study of five home use LED devices found three of the five shipped with essentially identical manuals, two lost around half their power output within three minutes of running, and effective treatment area varied from 7 square centimetres at contact to 154 at a short distance. The manufacturers' own dosing instructions were found to be inconsistent with delivering the dose they described.
The professional literature is not much better. The systematic review most often cited in this field notes that in many of its 45 studies the treatment parameters were not fully described, which is a large part of why the results conflict.
None of that makes the underlying effect imaginary. It makes the specification the only thing separating a device that might reproduce a published result from one that cannot be related to any of them.
Our guide to what the evidence supports by condition goes through where that published result is strong and where it is not.
Reading a Product Page Quickly
If the page gives you a wavelength in nanometres, an output in milliwatts and a laser class, you can do the arithmetic above and decide for yourself. If it gives you a colour, a wattage that turns out to be the wall adapter, and a testimonial, there is nothing to calculate.
That is the entire test, and it takes about thirty seconds.
The at home laser device we sell publishes 650 and 808 nanometres and its optical output for that reason.
Numbers are only half of it, though, because how you move the head over the dog decides the area and therefore the dose.
Our guide to running a session explains what to do with those figures in practice.
Numbers you can check are the only part of this category that is not a claim.
The Whisker Bark red light therapy device is built around that idea, which is also why it ships with safety glasses rather than a suggestion that you find some.
Share
