Blue-light filters and contrast sensitivity: what the data actually says

Walk into any optical shop in the last five years and a salesperson will, within minutes, offer to add a blue-light filter to your prescription. The pitch is well-rehearsed. Screens emit blue light. Blue light strains your eyes, ruins your sleep, and is slowly cooking your retina. For another fifty or hundred dollars, this coating will fix it.

It's a clean story, and an oversold one. The mechanism the marketing leans on is real — short-wavelength light does interact with non-image-forming pathways in the retina, including the cells that talk to your circadian clock. But the leap from "this wavelength does a thing in your retina" to "this spectacle coating will improve how you function at a screen" is a long one, and the research that's tried to bridge it has mostly come back with shrugs.

This post is for people who bought blue-light glasses and want to know if they actually do anything — and specifically, whether they do anything for the part of vision this site cares about: contrast sensitivity. We will go through the three claims separately (eye strain, sleep, retinal protection), summarise the Cochrane review, then close in on the CSF question. We will also draw the line between consumer blue-light glasses and FL-41 — a different tint with a different use case, often confused with the cheap blockers in the same display case.

What "blue light" actually means

The visible spectrum runs from roughly 380 nm at the violet end to 780 nm at the red. "Blue light" in the optical sense is the short-wavelength band — roughly 400 to 500 nm. Within that, there's a useful split:

• 400–440 nm — the high-energy visible (HEV) end, sometimes called "violet-blue." More energetic photons; the band marketing speak loves.
• 440–500 nm — blue-turquoise, including the ~480 nm peak that drives the melanopsin pigment in the retina's intrinsically photosensitive ganglion cells (ipRGCs). The circadian-relevant portion lives here.

Different "blue-light blockers" target different sub-ranges and vary widely in how aggressively they filter. A typical mass-market clear coating absorbs about 10–25% of light in the blue band; deeper amber or orange-tinted gaming/sleep glasses block 50–90% of it but visibly tint the world.

A piece of context worth carrying forward: the amount of blue light reaching your eyes from a phone, laptop, or tablet at typical brightness is a small fraction — on the order of a hundredth to a thousandth — of what you receive from natural daylight outdoors on an overcast afternoon. (That is an order-of-magnitude comparison, not a precise measured ratio, and it is ours rather than a Cochrane finding.) If consumer screens were dangerous blue-light sources, every outdoor walk would be much worse.

The three claims, separately

The blue-light pitch usually rolls three different claims into one product. The evidence base is uneven across them.

Eye strain reduction. The most-pitched, least-supported claim. The 2023 Cochrane systematic review (Singh et al., Cochrane Database of Systematic Reviews) included 17 randomised controlled trials from six countries and concluded that blue-light filtering spectacle lenses may make little or no difference to eye strain or visual fatigue with computer use over short-term follow-up, compared to non-filtering control lenses — the evidence on that specific outcome was rated low-certainty.¹ The earlier Lawrenson, Hull and Downie 2017 review in Ophthalmic & Physiological Optics — three eligible RCTs, 136 participants — reached the same conclusion with weaker evidence: no high-quality evidence to support the claim.² Two independent reviews, six years apart, the same answer.

Sleep improvement. The mechanism is firmer here. ipRGCs detect short-wavelength light and project to the suprachiasmatic nucleus, the brain's circadian clock; bright evening light suppresses melatonin and delays sleep onset. The Chang, Aeschbach, Duffy and Czeisler 2015 PNAS study showed four hours of pre-bedtime iPad reading suppressing evening melatonin by about 55% and delaying the circadian melatonin rhythm by more than 90 minutes versus paper-book reading.³ That is real circadian biology. The question is whether spectacle filters meaningfully intercept it. The Cochrane review's verdict on sleep outcomes was indeterminate — mixed results across heterogeneous populations, no consistent benefit.¹ Behaviour change (lower brightness, less evening screen time, dim ambient light an hour before bed) appears to do more for sleep than any spectacle coating tested.

Retinal protection. The strongest version of this claim — that consumer screens damage the retina and blue-blocking lenses prevent it — is not supported at consumer-screen brightness. The Cochrane reviewers found that no eligible trial had even measured macular or retinal outcomes, so there is no clinical evidence that blue-light filtering lenses protect macular health, and the light dose from screens is small relative to natural daylight.¹ A different, narrower question — whether yellow-tinted intraocular lenses inserted after cataract surgery modulate retinal short-wavelength exposure — has its own literature but is a surgical decision, not a consumer-optics one.

Stacked: eye-strain claim is unsupported, sleep claim has mechanistic grounding but spectacle filters specifically don't reliably deliver, retinal claim is not supported for consumer screens.

What about contrast sensitivity specifically?

A site about contrast sensitivity should be careful here, because the literature on blue-light filters and CSF specifically is small. Most RCTs in the Cochrane review measured eye strain, acuity, and sleep; far fewer measured the contrast sensitivity function directly. Of those that did, the picture is consistent.

Leung, Li and Kee's 2017 pseudo-randomised crossover trial — 80 computer users split between young-adult and middle-aged cohorts, three lens conditions including two blue-blockers and a clear control — measured contrast sensitivity using the Mars chart under both standard and glare conditions. They found no significant differences in contrast sensitivity between blue-light filtering and clear control lenses, in either age group, in either lighting condition.⁴ A separate six-month study by a different group, extending the wear period, reached the same no-difference conclusion for contrast perception.⁵

The theoretical reasons this null result is unsurprising:

• Short-wavelength input contributes little to luminance contrast at high spatial frequencies. The S-cones (blue-sensitive) are sparse — about 5–10% of cone receptors, almost entirely absent from the central fovea. The achromatic luminance pathway that carries fine-detail contrast is dominated by L- and M-cones. Cutting 10–25% of blue input from this signal makes almost no difference to mid- and high-frequency CSF.
• Veiling glare from blue scatter. A more interesting theoretical wrinkle: in eyes with significant intraocular scatter (advanced cataract, some corneal conditions), short-wavelength light scatters more than long-wavelength light. Filtering blue could in principle reduce veiling luminance and improve contrast in those specific eyes. Hammond and colleagues' work on macular pigment — the eye's own short-wavelength filter — found that people with denser macular pigment performed better on chromatic-contrast and glare-disability measures.⁶ But that's a story about the eye's own pigment, under specific glare conditions. Generalising to "blue-light spectacle coatings help everyone" is a leap the evidence does not support.

Net empirical effect on CSF in healthy adults at normal screen viewing: small, often statistically null, no consistent direction.

FL-41 is a different product

A frequent source of confusion: the rose-tinted FL-41 lens is not the same as a consumer blue-light coating, and the evidence behind it is different.

FL-41 was developed in the late 1980s to reduce sensitivity to fluorescent lighting. It is a broadband rose tint that attenuates the blue-green band around 480–520 nm — the spectral region most relevant to the ipRGC-driven non-image-forming pathway, and overlapping the cone-driven wavelengths implicated in migraine photophobia in more recent work. The most-cited early trial — Good, Taylor and Mortimer 1991 in Headache — compared a rose-coloured tint against a density-matched blue tint in 20 children with migraine over four months; only the rose-tint group sustained a drop in attacks, with mean frequency falling from about 6.2 to 1.6 per month.⁷ That study tested a generic rose tint, not FL-41 by name — FL-41 is a rose tint, which is why it is cited as the rationale, but the FL-41-specific migraine evidence is thinner and largely extrapolated from it. The broader evidence base has since extended to adult migraine, benign essential blepharospasm, and photophobic post-concussion patients.

Distinctions from consumer blue-light coatings:

• Different wavelength range. FL-41 cuts deeper into 480–520 nm rather than the 400–440 nm "HEV" band most clear blue-light coatings target. The pinkish-rose visible tint reflects this.
• Different population and indication. FL-41 is prescribed for photophobia and migraine, not general eye strain. Outcome measures are attack frequency and light-sensitivity scores, not acuity or contrast.
• No CSF-improvement claim. FL-41 doesn't claim to improve contrast sensitivity. It claims to reduce light-sensitivity-related symptoms in specific patient populations.

If you have migraine with significant photophobia, or post-concussion light sensitivity, FL-41 is a reasonable conversation to have with a neuro-ophthalmologist or headache specialist — the mechanism is covered in the photophobia post. What FL-41 isn't, despite sharing the lens-coatings menu, is a consumer-grade fix for screen-related eye fatigue.

What might actually help screen-related eye fatigue

The data is reasonably clear that blue-light coatings do not, on their own, fix what most patients describe walking into the shop. A few things in the same neighbourhood have better evidence or stronger mechanistic grounding:

• Adequate ambient lighting. Reading off a bright screen in a dim room is a high-contrast adaptation mismatch and a setup for fatigue. The task-lighting post walks through specific lux targets and fixture placement.
• Anti-reflective coatings. Internal reflections off the back surface of spectacle lenses and stray glare bouncing off screen surfaces add real veiling luminance. AR coatings cut this. Not glamorous; useful.
• Larger text and a correct prescription. A lot of what gets called "blue-light strain" is uncorrected presbyopia, undercorrected refractive error, or a screen sized for someone with sharper near vision than the user actually has. Bumping system text size one notch and confirming the prescription is current beats a coating every time.
• Breaks. The 20-20-20 rule (every 20 minutes, look at something 20 feet away for 20 seconds) has weak empirical evidence but a low-cost rationale. Probably doesn't hurt.
• Hydration and blink rate. Sustained screen use roughly halves blink rate. Dry-eye-driven discomfort is often misattributed to "blue-light fatigue" because it shows up around screens.

Where to spend money

A pragmatic ranking, for someone wondering what's worth paying for:

• Migraine with photophobia, or post-concussion light sensitivity: FL-41 is a real option worth raising with a specialist. Commonly $50–150 added to a prescription.
• Sleep concerns: built-in OS night-shift / warm-mode filters are free and directionally aligned with the mechanism. Lower screen brightness in the evening matters more than spectacle coatings.
• General "eye strain" alone: anti-reflective coating plus a lighting upgrade is a better investment than a blue-light coating. AR coatings have a long, boring track record of working.
• Blue-light coating as a standalone purchase: if the rest of your optics is already optimal (current prescription, AR coating, decent screen, adequate room lighting), the marginal benefit of a blue-light layer is probably indistinguishable from zero. Don't pay a premium.

This isn't advice to throw existing blue-light glasses away — the Cochrane review found no evidence they hurt either. It is advice not to let them be your first line of defence against problems that have better levers.

What our test can tell you

Note. A contrast sensitivity result is not a verdict on blue-light filters. It is a measurement of your photopic CSF under a specific calibrated condition.

If you want a personal answer to "are these glasses doing anything for me," the cleanest informal test is to take the same contrast measurement twice — once wearing the filter, once without — under the same lighting, on the same screen, at the same time of day, with the same recent sleep and caffeine. The published RCTs predict the difference will be small or absent. If you find a meaningful difference in your own readings, that's a data point worth taking seriously, but understand that single-session test-retest noise on this measurement is already about ±0.15 log units, so small differences are likely noise.⁸ A difference of ≥0.30 log between conditions, repeated across multiple sessions, would be the threshold to treat as more than chance.

Take the test

Take the test once with your glasses on and once with them off — or with and without any filter coating you're curious about. Match the conditions. Look at both numbers. The honest expectation, from the literature, is that they will be close. That's not a disappointment; that's information about what the coating is and isn't doing.

If you want background on what contrast sensitivity actually measures before running the comparison, the primer post is the place to start. The screen-calibration post covers why a calibrated baseline is the only condition under which the with-versus-without comparison is fair.