Skip to main content

​
Why do studies on blue light glasses disagree with each other?

By Spektrum Glasses Editorial Team · Published 2026-08-05 · Updated 2026-08-05 · Facts re-checked 2026-08-05 How this page is written and checked: our editorial method · how we verify claims

​
Short answer

Because they are not testing the same thing. “Blue light glasses” names an intention, not a specification: lenses sold under that label range from filtering roughly a third of 450 nm light to filtering over 97% across 400-500 nm. Add small samples, subjective endpoints and unblinded tinted lenses, and disagreement is the expected result.
  • The category name specifies no wavelength, no amount, and no test method.
  • Our clear lens filters 99.99% at 400 nm but 33.1% at 450 nm - same lens.
  • Leung 2017 (PLOS ONE) measured ~10-24% blue-light-hazard reduction in commercial filtering lenses.
  • A tinted lens cannot be blinded, so participants always know their group.
  • Demand a report number, an accrediting body, and the band every percentage covers.

​
The label covers lenses that differ by a factor of three

Two studies can both say they tested “blue light glasses” and be testing products with almost nothing in common. The phrase describes what a lens is meant to do, not what it measurably does. It carries no wavelength range, no minimum filtering level, no test standard, and no requirement that anyone measured the lens at all. Here is the spread across our own products, all measured, all with the band stated: The amber lens removes roughly twice as much of the 400-500 nm band as the ZENOX clear lens, and about three times what our own clear lens does at 450 nm. Leung, Li and Kee (PLOS ONE, 2017) measured commercially available blue-light-filtering lenses and found they reduced the calculated blue-light hazard by roughly 10-24% - a real published figure, and one that sits far below what a saturated amber lens does. So when a review article pools trials of both kinds of lens under one heading, the results disagree for a purely mechanical reason: the dose was different. That is not scientific controversy. That is a missing specification.

​
The size of a percentage comes from the band you pick

This is the arithmetic every reader should be able to run, including on us. Take our clear lens. At 400 nm it filters 99.99%. At 450 nm the same lens, in the same report, filters 33.1%. Both numbers are true. A vendor quoting only the first one has not falsified a measurement - it has chosen the most flattering single point on a curve that falls steeply from left to right. Even averaging is easy to inflate. The four published points for that lens are 99.99, 95.1, 63.0 and 33.1. A naive average is about 72.8%, but that is not a legitimate band figure either: it ignores the shape of the curve between the points and any weighting. A proper number states the range it integrates over, which is why we write “97.9% across 400-500 nm” rather than “blocks 98%.” There is a physical check that costs nothing. Our clear lens has 91.6% visible (photopic) transmission - it looks near-clear. Light in the 450-500 nm region is visible light. A lens cannot remove most of it and still look colorless. If a lens looks clear and is advertised as removing 90%+ across 400-500 nm, the claim is contradicted by the fact that you can see through it normally. Heavy filtering across the whole blue band shows up as visible color, every time.

​
Study designs differ as much as the lenses do

Even where lenses match, methods rarely do. A few sources of divergence are worth knowing by name.
  • Blinding. An amber lens is obviously tinted, so participants know which group they are in and expectation rides along with the result. A near-clear lens can be matched against a plausible sham; a strongly colored one cannot.
  • Endpoints. Many studies rely on self-reported questionnaire scores rather than an objective measurement. Questionnaire scales differ between studies, so results are not directly comparable.
  • Exposure. Screen brightness, ambient room light, task duration and viewing distance all change how much blue-wavelength light reaches the eye. Two protocols with identical lenses can deliver different doses.
  • Size and duration. Small participant counts and short study periods produce wide confidence intervals, and wide intervals are how honest studies end up pointing in opposite directions.
We are deliberately not citing a list of individual trials here. We hold ourselves to the sourcing rule we are asking you to apply to everyone else: we cite what we have read and can name, and otherwise we say the evidence is limited and contested rather than filling the gap with a reference that sounds authoritative.

​
What a defensible number looks like

A percentage is auditable when it travels with the information needed to check it. Ours reads: 97.9% across 400-500 nm, amber lens, COLTS Laboratories report O-SPG111015, spectral transmittance per ANSI Z80.3, lab accredited to ISO/IEC 17025 by A2LA under certificate 1612.01. Ask any vendor - us included - for the same five things:
  1. The band. Which wavelengths does the percentage cover? A single number with no range is unverifiable.
  2. The report. A document number and date, not a marketing graphic redrawn in brand colors.
  3. The lab. Accredited to ISO/IEC 17025, and by whom. A2LA is one accreditor; the point is that someone independent audited the lab’s competence.
  4. The method. ANSI Z80.3 is the transmittance standard for non-prescription eyewear. Naming a method means the test is repeatable.
  5. The exact item tested. Lens color, coating and model. A report for the amber lens says nothing about the clear one.
If a request for the report is met with a testimonial, a certificate with no issuing body, or a number with no band, you have learned what you needed to know without reading a single study.

​
Where this leaves a buyer

Two separate questions get tangled together, and separating them resolves most of the apparent contradiction. How much light of a given wavelength a lens removes is a settled, measurable engineering fact, testable in an accredited lab to a stated standard. Whether that reduction produces a noticeable difference for a particular person doing a particular task is a different question, and the published evidence for symptom benefit is limited and contested. A measurement cannot settle it, and we will not pretend our lab report does. What the measurements do tell you is which tool matches which intention. If you want substantial reduction across the whole blue band, a near-clear lens is the wrong instrument - 33.1% at 450 nm is what a near-clear lens can do while staying near-clear. Our amber evening lens reaches 98.3% at the 460-480 nm melatonin band and 97.9% across 400-500 nm, at the cost of obvious color. The orange lens (99.96% across 380-500 nm) and red lens (99.83% across the same band) go further still, and neither is suitable for driving. There are also changes to the light itself that cost nothing and are easy to test against: lowering screen brightness, reducing overhead lighting in the evening, and increasing text size so you sit further back. If you try a lens and notice nothing, that is a legitimate outcome and a common one. Our warranty runs 365 days; Amazon orders follow Amazon’s 30-day return window and orders from spektrumglasses.com have 90 days. The return window, not the marketing, is where you find out whether a lens does anything for you.

​
Where these numbers come from

Every measured figure quoted here is transcribed from a third-party laboratory report, published in full with the wavelength band and the report number: lab results. Our rule for what may appear on this page at all is on how we choose what to publish.