Blue light glasses: what a 2023 review of the sleep evidence actually found

Blue-light blocking glasses are sold on a claim that has two parts, and the marketing keeps them close enough together that they read as one fact instead of two. The first part is real. The second has been tested, and the testing did not go the way the product implies.

The physiology is not the disputed part

Evening light suppresses melatonin, delays circadian phase, and lengthens the time it takes to fall asleep — and blue-heavy light does more of this than other colours at equivalent brightness. That mechanism is described in NIH material on the body’s circadian clock: light reaching the eye tells the brain how much melatonin to release, and less melatonin in the evening means less sleepiness on schedule. Screens are blue-heavy relative to a lamp or a fire, and using one late at night will, on average, do something measurable to that signal.

None of that is in dispute. It is also not what is printed on the box. What is printed on the box is that this specific lens fixes it.

What the trials found

A 2023 systematic review pooled seventeen randomised trials of blue-light filtering spectacle lenses against clear or placebo lenses. On sleep quality, the reviewers rated the evidence low-certainty and the effect inconclusive — not “no effect,” but not the effect the marketing implies either, and not enough trials measuring the same outcome the same way to say much with confidence.

That is a specific and useful kind of negative result. It means the mechanism — light, melatonin, timing — is well established while the intervention built on top of it, this lens, worn for this many hours, filtering this much of the spectrum, has not cleared the bar. We have made this exact distinction before: a real cause is not evidence that a specific product addresses it, and the two get sold as though they were the same claim.

A mechanism being real is not evidence that a specific product corrects it. The review looked for the second part and did not find it.

Why weak trials look like this

Small samples, short exposure windows, mismatched lens tints, and self-reported sleep quality standing in for measured sleep are the usual reasons a category ends up with “inconclusive” instead of a clean answer either way — the same list of problems we’ve flagged in reading any accuracy figure: what was measured, on whom, and by what method changes the number more than most marketing acknowledges.

What we’d actually do

Keep the glasses if you already own a pair and like wearing them. No trial in the review found harm, and comfort at a screen in the evening is a reasonable thing to optimise for even without a sleep benefit attached to it.

Don’t buy a pair expecting a measurable difference in how fast you fall asleep or how well you sleep once you do. The trials that looked for exactly that mostly didn’t find it.

If sleep is the actual goal, the higher-value changes are duller and better supported: dim the room in the two hours before bed, put the phone somewhere other than arm’s reach, and let the light itself — not a lens sitting between you and it — do less in the evening. None of that has a price tag attached at the till, which is usually the first sign a claim is being sold rather than measured.