Lutein, Glare Recovery and Night Driving: What the Numbers From the Trials Show

Search this topic and you will read that carotenoids “greatly reduce the effects of glare” without a single effect size attached. The measured numbers are real, they are published, and they are more interesting than the marketing — because they include failures. Below are the three placebo-controlled trials that actually put a number on it, reported in full, including the endpoints that went nowhere.

The three trials that measured glare, side by side

These are the human, randomised, placebo-controlled studies that measured a glare-related or contrast-related endpoint after lutein supplementation. Doses, durations and results are as reported in the published abstracts.

MPOD = macular pigment optical density. “Significant” here means statistically significant versus placebo as reported by the authors, not clinically meaningful in daily life — a distinction none of these papers claims to have bridged.

What a photostress recovery test actually measures

Nothing on the first page of search results for this topic defines the test, which is odd, because the test is the whole reason “glare” and “glare recovery” produce different answers.

A photostress recovery test works like this. You first establish what a person can read — typically the smallest line of letters they can resolve, or the threshold at which they can detect a small target. Then you bleach the retina: a bright, brief, controlled flash of light aimed at the central macula, intense enough to temporarily deplete the photopigment in the cones there. For a few seconds afterwards, the person genuinely cannot see the target. The examiner starts a timer and stops it at the moment the original target becomes visible again. That interval — typically some tens of seconds — is the photostress recovery time. A shorter time means the eye is regenerating usable vision faster.

This is the closest laboratory analogue to being dazzled by oncoming headlights and then needing to see the lane markings again. It is not the same as the dazzle itself. And the mechanistic reason researchers expect macular pigment to matter here is straightforward: the yellow pigment layer formed by lutein and zeaxanthin absorbs short-wavelength light before it reaches the photoreceptors, so a denser layer means less bleaching from the same flash, and less to recover from.

Mesopic is the other term worth knowing, because it is where the Yao driver study found its strongest effect. Vision is described as photopic in bright light, when cones do essentially all the work; scotopic in near-darkness, when only rods respond; and mesopic in the range between, roughly 0.005 to 5 candela per square metre in most definitions, where rods and cones are both active at once. Street-lit night driving sits squarely in the mesopic band. That is why a result measured under mesopic conditions is more relevant to night driving than the same result measured under a bright test lamp — and why the testing conditions, not just the outcome, decide whether a study means anything to you.

Glare disability and glare recovery are two different measurements

This is the distinction every incumbent article flattens, and it is the single most useful thing on this page.

Glare disability asks: while a bright light source is in your field of view, how much worse does your vision get? It is measured with the glare source still on. Glare recovery asks: after the bright light is gone, how long until you see normally again? It is measured with the light off.

Hammond and colleagues measured both, in the same 115 people, on the same days, for a year. Recovery improved significantly. Disability did not. The authors did note that glare disability tracked macular pigment density across the study period — the relationship was there — but the supplemented group's glare disability did not rise significantly. Ma and colleagues, at 12 weeks, similarly reported no significant change in glare sensitivity while contrast sensitivity did move.

The one clear positive on a glare endpoint is Yao 2013, using a higher dose (20 mg lutein), a full year, and a driver population. That is a meaningful result. It is also a single trial, unreplicated at that dose, and it reported glare and contrast sensitivity together rather than separating disability from recovery. Two negatives and one positive on glare, against three positives on contrast or recovery, is the honest summary. Anyone writing “lutein reduces glare” without that split has not read the abstracts.

The one trial that was actually run in drivers

Yao and colleagues published the only study in this set designed around the occupational question rather than the laboratory one. One hundred and twenty healthy drivers, randomised, double-blind, placebo-controlled, 20 mg of lutein daily for a full year. Serum lutein rose. Central macular pigment optical density rose. Contrast sensitivity and glare sensitivity both improved significantly, with the authors specifically flagging the mesopic condition — the low-light state that matches real night driving — as where the effect was strongest. Vision-related quality of life improved on the driving subscale.

Three caveats belong in the same breath. First, best spectacle-corrected visual acuity showed no significant difference between groups, so this is not a study about seeing more sharply. Second, the dose is 20 mg, double the AREDS2 reference amount of 10 mg and higher than most consumer eye formulas carry; a result at 20 mg is not evidence for 10 mg. Third, the quality-of-life instrument is a self-report questionnaire in an unblinded-to-expectation population — participants knew they were in a lutein study, and subjective driving comfort is exactly the kind of endpoint that moves on expectation alone.

What the evidence does not show

Being specific about the limits is more useful than another round of enthusiasm.

It does not show improved eyesight. Visual acuity did not significantly change in Yao 2013 or Ma 2009. Contrast sensitivity is not acuity. You do not get a better line on the eye chart.

It does not show a consistent effect on glare itself. Two of three trials returned nulls on their glare endpoint. That is not a technicality; it is the majority of the evidence.

It does not show anything about safety on the road. Not one of these studies measured driving performance, reaction time behind a wheel, near-misses or collisions. Laboratory contrast charts are not a proxy for road safety, and we will not pretend otherwise.

It does not tell you about older eyes. Ma 2009 enrolled 22- to 30-year-olds. Hammond 2014 enrolled healthy young adults. The people who most often complain about headlight glare are over 60, where lens yellowing and light scatter inside the eye are major contributors that no carotenoid addresses.

If you are reading a label because of night driving

Three practical points follow from the trials rather than from marketing copy.

See an eye-care professional first. Worsening glare at night, halos around lights or a sense that oncoming headlights are getting harder to handle are symptoms an optometrist or ophthalmologist should look at. Early cataract is the most common cause and it is diagnosed in a chair, not solved by a capsule. The underlying measurement that changes first is explained in contrast sensitivity explained. Nothing in this article is medical advice, and no supplement diagnoses, treats, cures or prevents any condition.

Check the milligrams against the trial numbers. The reference amounts worth comparing a Supplement Facts panel to are the AREDS2 figures of 10 mg lutein and 2 mg zeaxanthin. If a panel hides those inside a proprietary blend, you cannot tell what you are getting. This is also the reason our Visivra review declines to quote milligram amounts we have not verified from the label itself.

Give it the duration the studies used. The shortest trial here ran 12 weeks and the two strongest ran a year. Judging a carotenoid after three weeks tells you nothing, because macular pigment accumulates over months. If you want the wider evidence picture on these two carotenoids, our review of the lutein and zeaxanthin evidence covers what the AREDS trials found and where the data stops. And if the formula you are holding also contains beta-carotene, read the trial doses behind the smoker warning before you buy. And if it leans on bilberry for its night-vision claim, that story is examined in what the bilberry trials actually found.

Frequently asked questions