What does the percentage on blue light glasses mean, and how do you know it is right?

What does the percentage actually measure?

A lens is measured in a spectrophotometer. The instrument sends light through the lens one wavelength at a time and reads how much comes out the other side. The result is a curve, the transmission curve, showing what passes at every nanometre from ultraviolet to red. The percentage on a product page is a summary of that curve over a chosen band.

That is where the figures part ways. The same lens can be described by its best value at a single wavelength, by an average across a band, or by a band so narrow that almost any lens reaches 99%. A clear lens can filter almost everything at 400 nm, where the light is close to invisible anyway, and almost nothing at 460 nm. A seller who writes "blocks 99% of blue light" without naming the band has not lied, but has not said anything either.

Our figures are stated per lens and per band, and they come from the supplier's spectral measurement of each lens tint. When a product page says "up to 99%", it means the highest-filtering lens sold with that particular frame, never a value we worked out ourselves.

Why does the wavelength band matter more than the percentage?

Because the eye does not care about all blue light equally. The retina has a cell type that forms no images but measures light and reports to the brain's clock. When Brainard and colleagues mapped its sensitivity in 2001, with 72 subjects and 627 melatonin measurements at wavelengths between 420 and 600 nm, the most potent region sat between 446 and 477 nm. The cell is described in Your eye's third cell type.

That is the middle of a screen's light. When researchers at King's College London and the University of Surrey measured an iPad Air, an iPhone 5s and a Kindle Paperwhite in 2015, all three had their blue peak between 445 and 455 nm. A lens that filters hard at 400 to 420 nm but passes everything at 450 nm has missed precisely the band the screen emits and the clock listens to. That is why our bands start at 400 nm and run to 500 nm for yellow and orange, and to 550 nm for red, which also takes part of the green.

The same goes for the word UV400. According to the WHO, ultraviolet radiation runs from 100 to 400 nm, with UVA between 315 and 400 nm. UV400 means the lens filters up to 400 nm, that is, up to where visible light begins. It says nothing about blue light between 400 and 500 nm. So when UV400 appears on a product page, including ours, it is a different fact from the blue light filtration.

How much do ordinary clear blue light glasses filter?

Less than you would think. The 2023 Cochrane review covered 17 randomised trials of blue-light filtering lenses, with between 5 and 156 participants per trial and follow-up from under a day to five weeks. Lead author Sumeer Singh of the University of Melbourne told Optometry Today that such lenses typically filter out about 10 to 25% of blue light, depending on the product, and that filtering more would require an obvious amber tint that affects colour perception.

That is the simplest check of all. A clear lens claimed to filter 90% of blue light across the whole band is at odds with physics. If a lens filters a lot of blue it looks yellow, orange or red, because blue is what is missing from the light that passes. The colour is not a style, it is the receipt. What Cochrane actually measured, and why evening lenses are a different product from office glasses, is in Do blue light glasses work?.

Which three questions expose an unmeasured figure?

• Which wavelength band does the percentage cover? The answer should be two numbers in nanometres, for example 400 to 500 nm. Without the band the figure cannot be compared with any other.
• Is it a peak value or an average, and who measured it? A serious figure comes from a spectral measurement per lens tint, done by the manufacturer or a laboratory, not from an estimate.
• How much visible light does the lens let through? A lens that is dark enough removes all blue by removing everything. What matters is how much blue disappears relative to how much light remains.

The third question has a scale of its own. The ISO 12312-1 standard sorts sun lenses into categories by how much visible light they transmit: category 0 above 80%, category 1 from 43 to 80%, category 2 from 18 to 43%, category 3 from 8 to 18% and category 4 from 3 to 8%. Under the same standard, as reproduced by the UK College of Optometrists, lenses transmitting less than 75% are not suitable for driving at twilight or at night. That is why an orange or red evening lens belongs on the sofa and at the desk, not behind the wheel.

The British research group above also showed how the two figures differ. When they placed orange-tinted glasses in front of an iPad, the light weighted by the clock cell's sensitivity fell from 302 to 65 melanopic lux, roughly 79%. Visible light fell only from 319 to 202 lux, roughly 37%. That is what a good evening lens should do: remove the blue and leave the rest.

The figures are the supplier's spectral measurement per lens tint. We have the visible light transmission for LOWLIGHT ONE and will publish it for the tinted lenses as soon as the measurement is in. Until then we say no more than we can show. How the three tints differ across the day is covered in Yellow, orange or red lens.

What does the law say about figures in marketing?

The Swedish Marketing Act is clear. Under section 10, a trader may not use incorrect claims or other representations that are misleading, and the section singles out claims about a product's quality and other distinguishing properties. A filtration percentage is such a claim. That is why a figure should be traceable to a measurement, and why we state the band and the lens every time.

How does BLUR-IT state its figures?

Per lens, per band, never as 100%. The figure on a product page is the highest-filtering lens for that frame, and each lens has its own band. We do not write that the glasses improve sleep, reduce tired eyes or protect the retina, because those are questions the research is still testing and a transmission curve cannot answer. What we can show is what reaches the eye through the lens. The rest is your decision, and the whole line is at LOWLIGHT.