← All supplements

AREDS2 Eye Formula

Studies disagree — unsettled

A commercial genetic test told people with one eye genotype to skip zinc and another to skip antioxidants. Three independent teams re-ran the numbers behind it and found the claim did not hold up — but a separate, real trial found something narrower that might.

What this supplement connects to

AREDS2 Eye Formula Variant: rs1061170 · CFH rs1061170 · CFH Variant Variant: rs10490924 · ARMS2 rs10490924 · ARMS2 Variant AREDS2 Eye Formula AREDS2 Eye Formula Antioxidant …

The AREDS2 formula (vitamin C, vitamin E, zinc, copper, and lutein/zeaxanthin in place of the beta-carotene the original 1994 AREDS formula used — beta-carotene was dropped after the trials on this site's antioxidants page found it raised lung cancer risk in smokers) is a standard recommendation for people with intermediate AMD, reducing progression to advanced disease by roughly a quarter over five years in the original AREDS trial.

A pharmacogenomic claim, and what happened to it

Awh et al. 2013 reanalyzed AREDS genetic data from 995 white participants and reported that CFH and ARMS2 genotype predicted who benefited from which part of the formula: people with CFH risk alleles seemed to do worse on zinc (RR=1.83, P=0.01, for two risk alleles), while people with ARMS2 risk alleles seemed to do worse on antioxidants (RR=2.58 to 3.96, P=1.04×10⁻⁶). The paper's own conclusion recommended matching supplement type to genotype — and a commercial genetic test was built and sold on that basis.

Assel et al. 2018, published in the same journal, had three independent statistical teams re-run the analysis on the AREDS data, including a comparison against the original investigators' own dataset. They found errors in the data underlying the 2013 claim, could not replicate any genotype-treatment interaction after correcting for multiple testing, and could not confirm the claim that supplementation raised AMD risk in any genotype subgroup. Their conclusion: patients should be offered the standard zinc and antioxidant formula without consideration of genotype.

A narrower, still-open question

Separately from the AREDS2-formula controversy, a real randomized trial asked a more specific question about one ingredient. The NAT2 study (Merle et al. 2015) randomized 250 people who already had neovascular AMD in one eye to daily fish-oil capsules (280 mg DHA, 90 mg EPA, 2 mg vitamin E) or placebo, and tracked whether their other eye — which still only had early disease — developed new vessels. Among people with the non-risk CFH genotype, DHA supplementation cut that occurrence from 38.2% to 16.7% (P=0.008), with a significant genotype-by-treatment interaction (P=0.01). People carrying a CFH risk allele saw no such benefit.

This is a genuinely different kind of finding from the AREDS2 controversy above: a prespecified randomized trial rather than a post-hoc reanalysis, on a single ingredient (DHA) rather than the whole formula. It has not been independently replicated, and it does not currently change any guideline — but it is the kind of result the field would need more of before genotype-guided eye supplementation becomes anything other than contested.

Sources

Papers, with their authors

Questions about AREDS2 Eye Formula

What is the AREDS2 Eye Formula?

AREDS2 is a specific vitamin and mineral formula (vitamin C, vitamin E, lutein, zeaxanthin, zinc and copper) shown in a large clinical trial to slow progression of intermediate age-related macular degeneration in some patients.

Should I get genetic testing before starting AREDS2 supplements?

No — the largest independent re-analysis of the data behind that recommendation found errors in it and could not replicate the claimed genotype-treatment interaction. Standard guidance is to take the zinc-and-antioxidant formula without genotype testing.

What was the original claim, and who debunked it?

Awh et al. 2013 claimed CFH genotype predicted a harmful response to zinc and ARMS2 genotype predicted a harmful response to antioxidants, and a commercial test was sold on that basis. Assel et al. 2018 had three independent statistical teams re-analyze the same data, found errors in it, and could not reproduce the interaction.

Is there anything to the genotype story at all?

One narrower finding, separate from the AREDS2 controversy: a real randomized trial (the NAT2 study) found that DHA fish-oil supplementation specifically benefited people with the non-risk CFH genotype, cutting new-vessel occurrence from 38.2% to 16.7%. It has not been independently replicated.

Why was beta-carotene removed from AREDS2?

Two earlier trials found that beta-carotene supplementation raised lung cancer risk in smokers and former smokers — covered on this site's antioxidants page. AREDS2 replaced it with lutein and zeaxanthin.

Nothing on this page recommends taking anything. If a result here matters to you, take it to a doctor or a pharmacist rather than to a shop.