Energy & Longevity
Lutein and Zeaxanthin for Eyes: 29% vs 31%, and P Was .12
Key takeaways
- AREDS2's primary randomised comparison of lutein + zeaxanthin against placebo was not statistically significant: 29% vs 31% progression at five years, P = .12.
- Omega-3 (DHA 350 mg + EPA 650 mg) had no effect at all — HR 0.97, P = .70 in the trial, and HR 1.01, P = .91 at ten years.
- The formula that does have randomised evidence is the original 2001 AREDS combination, and only the combination: antioxidants alone (OR 0.80) and zinc alone (OR 0.75) both failed to reach significance on their own.
- Everyone studied already had intermediate AMD. Participants in the lowest-risk category had only a 1.3% five-year chance of progressing, so there was almost nothing for a supplement to prevent.
- Beta carotene was dropped for a reason: 23 lung cancers versus 11, mostly in former smokers, and an odds ratio of 1.82 at ten years.
Walk down the supplement aisle and lutein is on every eye product in it. The reason given is always the same trial: AREDS2, run by the National Eye Institute, 4,203 participants, five years of follow-up, published in JAMA in 2013. It is a real trial, and it is large, and it is well conducted. It is also a trial whose primary result was negative, which is the part that does not make it onto the label.
The number the labels skip
AREDS2 randomised participants aged 50 to 85 who already had age-related macular degeneration — bilateral large drusen, or large drusen in one eye with advanced AMD in the other — to lutein 10 mg plus zeaxanthin 2 mg, to DHA 350 mg plus EPA 650 mg, to both, or to placebo, all on top of the original AREDS formula. The outcome was progression to advanced AMD, counted by eye.
By five years, the Kaplan-Meier probabilities were 31% for placebo and 29% for lutein plus zeaxanthin. Two percentage points. The hazard ratio was 0.90 with a 98.7% confidence interval running from 0.76 to 1.07, and P = .12.
A confidence interval that crosses 1.0 and a P value of .12 mean the trial could not distinguish that difference from chance. The authors did not soften it. Their conclusion reads: addition of lutein + zeaxanthin, DHA + EPA, or both to the AREDS formulation in primary analyses “did not further reduce risk of progression to advanced AMD.”
That sentence is the finding. Everything sold on top of it is built from what came afterwards.
Omega-3 was flatter still
The same trial tested fish oil, and here there is not even a two-point gap to argue about. Progression at five years was 31% on DHA + EPA against 31% on placebo: hazard ratio 0.97, 98.7% CI 0.82 to 1.16, P = .70.
The ten-year follow-up rechecked it and got HR 1.01, 95% CI 0.93 to 1.09, P = .91 — a result that sits almost exactly on the line of no effect. Omega-3 has arguments to make in other territory, and our look at who actually benefits from fish oil goes through where those arguments hold. Macular degeneration is not one of them. It was tested properly, at a real dose, in the right population, and it did nothing.
So why is lutein in the formula at all?
Because of beta carotene, and this is where the honest case for lutein actually lives.
The original 2001 AREDS formula contained beta carotene 15 mg. In AREDS2 the researchers also randomised participants to formulations with and without it, and the safety data came back badly: 23 lung cancers (2.0%) in the beta carotene group against 11 (0.9%) without it, nominal P = .04, mostly in former smokers. The ten-year follow-up confirmed it rather than dissolving it — odds ratio 1.82, 95% CI 1.06 to 3.12, P = .02. Beta carotene “nearly doubled the risk of lung cancer”, in the follow-up authors’ own words.
Beta carotene had to go. Lutein and zeaxanthin were the replacement, and they carried no comparable signal: OR 1.15, 95% CI 0.79 to 1.66, P = .46. When the ten-year analysis compared lutein/zeaxanthin directly against beta carotene, the hazard ratio for late AMD was 0.85 (95% CI 0.73 to 0.98, P = .02).
So the defensible claim is: lutein is a better carotenoid than the one we removed for causing cancer. That is a real finding and a genuinely useful one. It is not the same claim as “lutein protects your eyes”, and the distance between the two is the distance between P = .02 against beta carotene and P = .12 against placebo.
The formula that does have randomised evidence — and its own catch
Strip AREDS2 away and go back to the 2001 trial. 3,640 participants, average follow-up 6.3 years, four arms: antioxidants, zinc, antioxidants plus zinc, or placebo.
Antioxidants plus zinc reduced the odds of progressing to advanced AMD, odds ratio 0.72 with a 99% confidence interval of 0.52 to 0.98. That is the number the whole category rests on, and it holds up.
What is left out of the retelling is that neither half worked alone. Zinc by itself: OR 0.75, 99% CI 0.55 to 1.03. Antioxidants by themselves: OR 0.80, 99% CI 0.59 to 1.09. Both intervals cross 1.0. If you are taking a zinc supplement, or a vitamin C and E combination, and treating either as the eye-protective part, the trial you are citing tested exactly that and could not show it. The result belongs to the combination, at those doses, together. Our page on zinc covers where zinc’s own evidence is stronger, and this is not it.
Who was actually in these trials
This is the part that decides whether any of it applies to you, and it is the part almost never stated.
Every AREDS2 participant already had macular degeneration. Enrolment required bilateral large drusen, or large drusen in one eye and advanced AMD in the fellow eye. The original AREDS cast a wider net, and in doing so it produced the most clarifying number in either paper: participants with extensive small drusen, non-extensive intermediate drusen or pigment abnormalities had only a 1.3% five-year probability of progression to advanced AMD.
When the researchers removed those 1,063 lower-risk participants from the analysis, every effect estimate strengthened — antioxidants plus zinc went from OR 0.72 to 0.66. That is not a trick of statistics. It is what happens when you stop counting people who had almost no risk to remove.
The AREDS authors put the practical instruction in their own conclusion, and it is not a shopping instruction: persons older than 55 “should have dilated eye examinations to determine their risk of developing advanced AMD,” and those found to have extensive intermediate drusen, at least one large druse, non-central geographic atrophy, or advanced AMD in one eye “should consider taking a supplement of antioxidants plus zinc such as that used in this study.”
The eye exam comes first in that sentence. It comes first because without it you do not know which of the two groups you are in, and the trial only has an answer for one of them.
Where that leaves the bottle on the shelf
Three separate claims get blended together on eye supplement packaging, and they have wildly different evidence behind them.
The AREDS combination slows progression in people who already have intermediate AMD. Randomised, replicated, OR 0.72. Solid.
Lutein and zeaxanthin are the right carotenoid to use in that combination. Supported, mostly as a safety substitution and by direct comparison against beta carotene. Reasonable.
Lutein and zeaxanthin protect healthy eyes, or help with screens. Not tested in either trial. The primary randomised comparison against placebo, in people who did have disease, returned P = .12. In people without disease it has not been asked.
If your eyes are healthy and you are buying lutein as insurance, the trial on the box was run on someone else, and its main result did not reach significance even for them.
A note on the disclosure
The ten-year follow-up carries a conflict-of-interest statement worth quoting rather than skipping: the National Institutes of Health “holds a royalty-bearing license issued to Bausch and Lomb for the Age-Related Eye Disease Study (AREDS) supplement which ended in 2021”, and Bausch and Lomb “provided AREDS2 supplements for the AREDS2 study participants during the last 5 years of the epidemiologic follow-up study.”
The trials remain the best evidence in this area by a wide margin, and the disclosure is published openly, which is how it should work. But when a follow-up analysis produces the favourable lutein numbers that the randomised primary did not, it is worth knowing who supplied the product during that follow-up. Read the confidence intervals, note who paid, and weight the randomised result above the observational one — which in this case means weighting P = .12 above P = .02.
The bottom line
Lutein and zeaxanthin earned their place in eye formulas by being safer than beta carotene, not by beating placebo. In the trial everyone cites, 29% of lutein eyes and 31% of placebo eyes progressed to advanced macular degeneration, and the difference did not reach significance. Omega-3 was tested in the same trial and produced nothing at all. The combination that does work, works in people who already have the disease and who found out they had it from a dilated eye exam.
This article is information, not a diagnosis or a treatment plan. Age-related macular degeneration is diagnosed and staged by an eye doctor, and high-dose zinc and vitamin E have their own interactions and cautions. Talk to your ophthalmologist or optometrist before starting any AREDS-style formula.
Frequently asked questions
Does lutein and zeaxanthin actually work for eyes?
For the one group it was tested in — people who already have intermediate age-related macular degeneration — the honest answer is that the randomised primary result said no. AREDS2 gave 4,203 participants lutein 10 mg plus zeaxanthin 2 mg or placebo on top of the original AREDS formula, followed them for a median of five years, and found 29% of lutein eyes progressing to advanced AMD against 31% of placebo eyes. That is a hazard ratio of 0.90 with a 98.7% confidence interval of 0.76 to 1.07 and P = .12, which does not clear significance. The trial's own conclusion is that adding lutein and zeaxanthin 'did not further reduce risk of progression to advanced AMD'. Lutein's place in the modern formula was earned as a substitute for beta carotene, which is a different claim from working on its own.
Then why does every eye supplement contain lutein?
Because of what happened after the primary analysis. Beta carotene, the carotenoid in the original 2001 formula, produced 23 lung cancers against 11 in AREDS2, mostly among former smokers, and by ten years carried an odds ratio of 1.82 (95% CI 1.06 to 3.12). Something had to replace it, and lutein plus zeaxanthin was the candidate with no comparable safety signal (OR 1.15, P = .46). In the ten-year follow-up, lutein/zeaxanthin compared directly against beta carotene gave a hazard ratio for late AMD of 0.85 (P = .02). So the case for lutein is genuinely 'better than the ingredient we had to remove', not 'beats placebo'. Both things are true at once, and only one of them is printed on the bottle.
Should I take an AREDS2 formula if my eyes are healthy?
There is no trial evidence for that, because that person was never enrolled. AREDS2 required bilateral large drusen, or large drusen in one eye with advanced AMD in the other. The original AREDS was broader, and it reported something worth knowing: participants with extensive small drusen, non-extensive intermediate drusen or pigment abnormalities had only a 1.3% five-year probability of progressing to advanced AMD. When the researchers excluded those 1,063 low-risk people, the effect estimates got stronger — which is another way of saying the low-risk group was diluting a benefit they were never going to receive. The prerequisite is a dilated eye exam, not a supplement order.
What about lutein for screen fatigue and blue light?
That is a separate literature from AREDS and a much thinner one — small trials with subjective endpoints, often industry-funded, on branded extracts. Nothing in AREDS or AREDS2 speaks to it, because neither trial measured eye strain, sleep or screen comfort in the first place. If someone cites AREDS2 to you while selling a screen-fatigue product, they are citing a macular degeneration trial in people aged 50 to 85 with drusen. Treat the two bodies of evidence separately and judge the screen-fatigue one on its own small trials, which is a much less flattering exercise.
Did omega-3 do anything in AREDS2?
No, and it is the cleanest null in the whole trial. Participants received DHA 350 mg plus EPA 650 mg or placebo. Progression to advanced AMD at five years was 31% in the omega-3 arm and 31% on placebo: hazard ratio 0.97, 98.7% CI 0.82 to 1.16, P = .70. The ten-year follow-up was, if anything, flatter — HR 1.01, 95% CI 0.93 to 1.09, P = .91. That is about as close to zero as a large trial gets. Fish oil has arguments elsewhere; for macular degeneration this one was tested properly and it failed.
What exactly is in the formula that does have evidence?
The original AREDS combination from the 2001 trial: vitamin C 500 mg, vitamin E 400 IU, beta carotene 15 mg, zinc 80 mg as zinc oxide and copper 2 mg as cupric oxide. In that trial, antioxidants plus zinc gave an odds ratio of 0.72 (99% CI 0.52 to 0.98) for progression to advanced AMD. What is usually left out is that the components failed separately: zinc alone was 0.75 (99% CI 0.55 to 1.03) and antioxidants alone were 0.80 (99% CI 0.59 to 1.09), neither significant. The modern AREDS2 version keeps vitamins C and E, zinc and copper, drops beta carotene, and puts lutein 10 mg and zeaxanthin 2 mg in its place.
Is there a conflict of interest worth knowing about?
There is one, and it is disclosed in the ten-year paper rather than hidden: the National Institutes of Health held a royalty-bearing licence issued to Bausch and Lomb for the AREDS supplement, which ended in 2021, and Bausch and Lomb supplied the AREDS2 supplements to participants during the last five years of the follow-up study. That does not make the numbers wrong — they are published with confidence intervals and P values you can check. It does mean the ten-year follow-up is not a fully independent evaluation of the product it evaluates, which is the sort of thing worth reading before deciding how much weight the follow-up should carry against the randomised primary.
Can you get lutein and zeaxanthin from food instead?
Yes, and dietary intake is where the whole hypothesis came from — AREDS2's own stated rationale was that 'observational data suggest that increased dietary intake of lutein + zeaxanthin... might further reduce this risk'. The trial then tested that observation and the primary analysis did not confirm it. That sequence is worth holding onto, because it is the standard shape of a supplement story: an observational association, a trial that fails to reproduce it, and a product that keeps the observational claim on the label. Kale, spinach, and egg yolks remain the dietary sources people mean, and eating them costs nothing extra.
Sources
Every claim above is drawn from these primary sources. Last checked July 2026.
- 1.Age-Related Eye Disease Study 2 (AREDS2) Research Group. Lutein + zeaxanthin and omega-3 fatty acids for age-related macular degeneration: the AREDS2 randomized clinical trial. JAMA. 2013;309(19):2005-15. PMID 23644932. Interventions: lutein 10 mg + zeaxanthin 2 mg, or DHA 350 mg + EPA 650 mg. Primary outcome: HR 0.90 (98.7% CI 0.76-1.07), P = .12 for lutein + zeaxanthin; HR 0.97, P = .70 for DHA + EPA; lung cancers 23 (2.0%) vs 11 (0.9%), nominal P = .04. Checked 6 September 2026.
- 2.Age-Related Eye Disease Study Research Group. A randomized, placebo-controlled clinical trial of high-dose supplementation with vitamins C and E, beta carotene, and zinc for age-related macular degeneration and vision loss: AREDS report no. 8. Arch Ophthalmol. 2001;119(10):1417-36. PMID 11594942. n=3,640, mean follow-up 6.3 years; antioxidants plus zinc OR 0.72 (99% CI 0.52-0.98); zinc alone 0.75; antioxidants alone 0.80; low-risk group 1.3% five-year progression. Checked 6 September 2026.
- 3.Chew EY, Clemons TE, Agrón E, et al. Long-term outcomes of adding lutein/zeaxanthin and omega-3 fatty acids to the AREDS supplements on age-related macular degeneration progression: AREDS2 report 28. JAMA Ophthalmol. 2022;140(7):692-698. PMID 35653117. n=3,882 at ten years; beta carotene lung cancer OR 1.82 (95% CI 1.06-3.12); lutein/zeaxanthin OR 1.15 (P = .46); late AMD HR 0.91 (P = .02) vs no lutein/zeaxanthin and 0.85 (P = .02) vs beta carotene; omega-3 HR 1.01 (P = .91). Includes the NIH/Bausch and Lomb royalty disclosure quoted above. Checked 6 September 2026.
- 4.National Eye Institute, National Institutes of Health — About AREDS and AREDS2, the agency that ran both trials describing their design and findings. Checked 6 September 2026.