Take-home points

  • Recent analyses of the AREDS1 and -2 datasets measured rates of geographic atrophy progression in each eye and related them to the randomizations in the clinical trials for robust assessment of treatment effects.

  • Oral micronutrient supplementation slowed GA progression towards the central macula, likely by augmenting the natural phenomenon of foveal sparing.

  • For extrafoveal GA, the AREDS2 supplements slowed the rate of progression to the fovea by approximately half; this was accompanied by supportive results according to GA area and visual acuity, particularly in relevant GA subgroups.

  • Given their excellent safety profile, convenient administration and low cost, patients with extrafoveal GA may benefit from AREDS2 oral supplements.

  • For slower progression, all patients with GA may also benefit from a healthy lifestyle, particularly a Mediterranean-type diet rich in vegetables and fruit, with less red meat, and avoiding cigarette smoking and heavy alcohol consumption.

  • Since these were post hoc analyses, we’re planning a new dedicated randomized clinical trial of oral supplementation for extrafoveal GA, as part of the AREDS3 program.

Bios

Dr. Hébert is a medical retina fellow at the National Eye Institute, National Institutes of Health in Bethesda, Md.
Dr. Indurkar is a medical retina fellow at the NEI/NIH.
Dr. Keenan is with the Division of Epidemiology and Clinical Applications, NEI/NIH.

Disclaimer
This research was supported by the Intramural Research Program of the National Institutes of Health (NIH). The contributions of the NIH authors were made as part of their official duties as NIH federal employees, are in compliance with agency policy requirements, and are considered Works of the United States Government. However, the findings and conclusions presented in this paper are those of the authors and don't necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.

AREDS2 oral supplements are well known to decrease the risk of progression to advanced age-related macular degeneration, particularly neovascular AMD.1-3 They’re recommended by the American Academy of Ophthalmology for patients with intermediate or advanced AMD in at least one eye.4 We hadn’t previously analyzed in detail whether they might also be helpful in slowing GA enlargement. We recently had the opportunity to address this important question, with the availability of GA data from reading center grading of all AREDS1/2 participants.5 Specifically, we calculated the GA enlargement rate for all eyes in these two large datasets with long follow-up time. Importantly, we can relate these enlargement rates to the underlying randomizations in the trials. Because of these robust randomizations, we’re justified in inferring causality and treatments effects, which isn’t true for observational data.

We used GA area as one outcome measure, in line with many recent clinical trials. However, for extrafoveal GA, the area metric has the problem that it treats all macular locations as equally important, whereas we know that the fovea and nearby paracentral areas are most important for visual acuity and quality of life.6,7 We therefore used GA proximity to the foveal center-point as a second outcome measure (Figure 1). For eyes with extrafoveal GA, time-to-fovea is increasingly recognized as an important metric,8 and slower change in proximity over time can indicate greater tendency to foveal sparing. Foveal sparing is an important protective phenomenon, whereby GA expansion near and into the fovea is usually much slower than expansion elsewhere, leading to beneficial GA configurations (e.g., horseshoe and donut) where the GA wraps around the fovea but doesn’t involve it until much later (Figure 1).9 These configurations are usually compatible with good visual acuity.10 Therefore, strategies to enhance foveal sparing would be very valuable, as they could keep eyes in these beneficial GA configurations for longer.

 

AREDS results

The AREDS randomization is shown in Figure 2A. In eyes with extrafoveal GA, for rate of progression towards the fovea, the oral antioxidant component (comprising vitamin C, vitamin E, and beta-carotene) had a significant treatment effect (Table 1).5 Participants randomly assigned to the antioxidants had a progression rate that was 36 percent slower than those randomly assigned to no antioxidants. Interestingly, no significant treatment effect was detected for area-based progression overall, although a beneficial treatment effect was present for the subgroup of eyes with early GA (particularly those with extrafoveal GA). In fact, in that subgroup of eyes with early extrafoveal GA, the rate of visual acuity decline was twice as slow in participants randomized to antioxidants vs no antioxidants. The zinc component had no significant beneficial or harmful effect for either the area or proximity outcome measure.

Figure 1. Color fundus photographs showing the progression of extrafoveal geographic atrophy towards the fovea over time. Geographic atrophy proximity (i.e., the shortest distance between the foveal center-point and the nearest pixel of GA) decreased gradually over time from 685 mm (2007) to 68 mm (2011). (This figure was published in: Keenan TDL, et al. Oral Antioxidant and Lutein/Zeaxanthin Supplements Slow Geographic Atrophy Progression to the Fovea in Age-Related Macular Degeneration. Ophthalmology 2025;132:1:14-29. Copyright Elsevier (2025))

AREDS2 results

Almost all AREDS2 participants were also assigned to take the AREDS supplements, so AREDS2 participants with extrafoveal GA were presumably already benefiting from the antioxidant component taken by all participants. In these analyses of the AREDS2 randomization (Figure 2B), we were therefore looking for any additional treatment effect. Interestingly, we saw a similar pattern of results to AREDS. In eyes with extrafoveal GA, for rate of progression towards the fovea, we did observe an additional significant treatment effect for the lutein/zeaxanthin component (Table 2).5 Participants randomly assigned to the lutein/zeaxanthin component (and to no beta-carotene) had a progression rate that was 35 percent slower than those randomly assigned to no lutein/zeaxanthin. This was accompanied by a borderline significant difference for visual acuity, with slower decline in acuity in those assigned to lutein/zeaxanthin. Again, no significant treatment effect was detected for area-based progression. The docosahexaenoic acid/eicosapentaenoic acid component had no significant beneficial or harmful effect for either the area or proximity outcome measure.

Figure 2. Randomization schemes. A) Age-Related Eye Disease Study: Participants were randomized 1:1:1:1, using a 2x2 factorial design, to oral antioxidants (vitamin C, vitamin E, and β-carotene) only, zinc only, both antioxidants and zinc, or placebo. B) AREDS2: Participants were randomized, 1:1:1:1, using a 2x2 factorial design, to oral lutein/zeaxanthin only, docosahexaenoic acid (DHA) plus eicosapentaenoic acid only, both lutein/zeaxanthin and DHA/EPA, or placebo. All participants were also offered the original AREDS formulation to take alongside the randomly assigned primary treatment.
 

Putting together the results from both trials

The combined results suggest that the modern AREDS2 supplement (which includes the same vitamin C, vitamin E, and lutein/zeaxanthin doses) should slow GA progression towards the fovea by as much as half. Therefore, patients with extrafoveal GA should benefit from AREDS2 supplements (Figure 3). The supplements may substantially slow GA expansion towards and into the fovea, presumably by enhancing foveal sparing. Together with their oral administration, high convenience, low cost and excellent safety profile, this makes the supplements an attractive option, with very few downsides. Importantly, approximately two-thirds of cases arise as extrafoveal GA,11 so this is a common form of disease.

However, as these results come from post hoc analyses, we need to replicate them in a new dedicated trial. We’re therefore planning a prospective trial of oral supplements for extrafoveal GA, as part of the AREDS3, and hope to begin enrolling participants for this in the near future. In the meantime, the results represent the only large-scale data available examining GA progression according to oral supplement use by randomization.

Table 1. Geographic atrophy proximity-based progression and area-based progression in the Age-Related Eye Diseases Study (AREDS), according to randomized assignment.

Bold font: P-value significant at the 0.05 threshold
 

Biological plausibility and a unifying explanation

Multiple points support the idea of a genuine treatment effect. First, not all post hoc analyses are created equal. These results come from randomized data, so they represent a high level of evidence quality. By contrast, observational data on supplement use, even from clinical trials, suffer from high risk of confounding and poor ascertainment of actual supplement use; it’s therefore impossible to make claims of efficacy or lack of efficacy from them. Second, a similar pattern of results was seen independently in AREDS and AREDS2, which is unlikely to occur by chance. 

Third, the results are biologically plausible. For example, lutein/zeaxanthin are abundant in the central macula (given sufficient dietary or supplement intake, since these important carotenoids can’t be synthesized by the body) but virtually absent in the peripheral macula.12 Therefore, it seems likely that, if their presence and antioxidant activity are able to slow GA progression, this effect would be observed preferentially at the central macula, where they're located. This effect would be captured well by the proximity measure but very poorly by the GA area measure, which is dominated by the more peripheral macula. These ideas are also supported by the subgroup analyses.

Fourth, the visual acuity data provide some support for the structural data. We had minimal expectation that slowing GA progression to the fovea would be accompanied by slower decline in acuity, since progression fell well short of actual center-point involvement in most cases. Despite this, significant or numerical differences were seen for some important comparisons.

Finally, in separate analyses of the same datasets according to dietary intake, as opposed to supplement use, we found similar associations, with strong associations between healthier diet and slower GA progression.13-15 In this way, both observational data on dietary intake and randomized data on oral supplementation point to the importance of antioxidants in slowing GA progression. In fact, in analyses that considered diet and randomized supplement intake together, the results for extrafoveal GA even demonstrated a degree of redundancy. With increasing oral supplementation by randomization, the associations between a healthier diet and slower progression to the fovea were fewer and weaker.15 It's very difficult to explain these results in the absence of a genuine treatment effect of oral supplements.

One potential limitation of these analyses is the use of color fundus photography, which was the gold standard for grading GA area and proximity at the time of the trials. However, levels of inter-grader agreement for measuring GA area on the AREDS2 CFP images have been analyzed in detail, and had low mean difference, narrow limits of agreement and no systematic bias.16 Although fundus autofluorescence and/or OCT imaging are now typically used, several studies have shown high levels of correlation between CFP and FAF images in measuring GA area and progression.17-19 In previous large-scale analyses of the AREDS2 dataset itself, measurements of GA progression rates were extremely similar between the two modalities, with no significant difference.17 In fact, the evaluation of foveal involvement was considered superior on CFP than on FAF. The central masked reading center grading and the repeated-measures regression (based on measurements at multiple time-points per eye) also mitigate this concern, through increased precision and power. Most importantly, even if the use of CFP were associated with slightly lower inter-grader agreement for GA measurements, this would apply to all randomized treatment arms. The randomization is key here and means that the results can’t be explained by the grading approach.

Figure 3. Diagram illustrating the potential benefits of oral micronutrient supplementation for different stages of age-related macular degeneration.

 

Lifestyle factors and geographic atrophy progression

More broadly, outside oral supplement use, our analyses strongly suggest that lifestyle factors are important not just for risk of GA occurrence but also for rate of GA progression.14 Analyses of the AREDS1/2 datasets have shown consistently that a Mediterranean-type diet is associated with slower GA area-based progression and slower progression to the fovea; for extrafoveal GA, this is accompanied by a slower decline in visual acuity.13,15 The most important components include vegetables, fruit, and less red meat. Cigarette smoking is associated with faster GA area-based progression, and, in men, heavy alcohol consumption is also associated with faster area-based progression.11,20 Overall, patients with GA may therefore benefit from both a healthy lifestyle and oral micronutrient supplementation.

Table 2. Geographic atrophy proximity-based progression and area-based progression in the Age-Related Eye Disease Study 2 (AREDS2), according to randomized assignment.

* All participants also assigned to receive AREDS supplements
† Considering only those participants in the secondary randomization study population who were randomly assigned to no β-carotene (analyzed because lutein/zeaxanthin and β-carotene compete for intestinal absorption)
DHA = docosahexaenoic acid; EPA = eicosapentaenoic acid
Bold font: P-value significant at the 0.05 threshold

 

Bottom line

Patients with extrafoveal GA should be counselled on the potential benefit of AREDS2 supplements in slowing GA progression to the fovea. In analyses that related GA progression rates back to the clinical trial randomizations, the AREDS2 supplements appear to slow progression to the fovea by approximately half, likely by augmenting the natural protective phenomenon of foveal sparing. Slowing the GA time-to-fovea in this way should help preserve visual acuity and quality of life. Patients with GA should also be advised on the likely importance of lifestyle factors in slowing progression, particularly adopting a healthy diet and avoiding smoking. RS

 

REFERENCES

1. Age-Related Eye Disease Study Research G. 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;119:1417-36.
2. Age-Related Eye Disease Study 2 Research G. Lutein + zeaxanthin and omega-3 fatty acids for age-related macular degeneration: The Age-Related Eye Disease Study 2 (AREDS2) randomized clinical trial. JAMA 2013;309:2005-15.
3. Age-Related Eye Disease Study 2 Research G, Chew EY, Clemons TE, et al. Secondary analyses of the effects of lutein/zeaxanthin on age-related macular degeneration progression: AREDS2 report no. 3. JAMA Ophthalmol 2014;132:142-9.
4. Vemulakonda GA, Bailey ST, Kim SJ, et al. Age-related macular degeneration preferred practice pattern. Ophthalmology 2025;132:P1-P74.
5. Keenan TDL, Agron E, Keane PA, et al. Oral antioxidant and lutein/zeaxanthin supplements slow geographic atrophy progression to the fovea in age-related macular degeneration. Ophthalmology 2025;132:1:14-29. .
6. Schmitz-Valckenberg S, Nadal J, Fimmers R, et al. Modeling visual acuity in geographic atrophy secondary to age-related macular degeneration. Ophthalmologica 2016;235:215-24.
7. Kunzel SH, Broadbent E, Moller PT, et al. Association of lesion location and functional parameters with vision-related quality of life in geographic atrophy secondary to age-related macular degeneration. Ophthalmol Retina 2024;8:794-803.
8. Zhang C, Kahan E, Begaj T, et al. Geographic atrophy natural history versus treatment: Time to fovea. Ophthalmic Surg Lasers Imaging Retina 2024;55:576-585.
9. Shen LL, Sun M, Ahluwalia A, et al. Natural history of central sparing in geographic atrophy secondary to non-exudative age-related macular degeneration. Br J Ophthalmol 2022;106:689-695.
10. Sunness JS, Rubin GS, Zuckerbrod A, Applegate CA. Foveal-sparing scotomas in advanced dry age-related macular degeneration. J Vis Impair Blind 2008;102:600-610.
11. Keenan TD, Agron E, Domalpally A, et al. Progression of geographic atrophy in age-related macular degeneration: AREDS2 report number 16. Ophthalmology 2018;125:1913-1928.
12. Owsley C, Swain TA, McGwin G Jr, Clark ME, Kar D, Curcio CA. Biologically guided optimization of test target location for rod-mediated dark adaptation in age-related macular degeneration: Alabama Study on Early Age-related Macular Degeneration 2 baseline. Ophthalmol Sci 2023;3:100274.
13. Agron E, Mares J, Chew EY, Keenan TDL, Group AR. Adherence to a Mediterranean diet and geographic atrophy enlargement rate: Age-Related Eye Disease Study 2 report 29. Ophthalmol Retina 2022;6:762-770.
14. Keenan TDL. Geographic atrophy in age-related macular degeneration: A tale of two stages. Ophthalmol Sci 2023;3:100306.
15. Agron E, Vance E, Domalpally A, Chew EY, Keenan TDL. Relationships between diet and geographic atrophy progression in the Age-Related Eye Diseases Studies 1 and 2. Nutrients 2025;17.
16. Danis RP, Domalpally A, Chew EY, et al. Methods and reproducibility of grading optimized digital color fundus photographs in the Age-Related Eye Disease Study 2 (AREDS2 report number 2). Invest Ophthalmol Vis Sci 2013;54:4548-54.
17. Domalpally A, Danis R, Agron E, et al. Evaluation of geographic atrophy from color photographs and fundus autofluorescence images: Age-Related Eye Disease Study 2 report number 11. Ophthalmology 2016;123:2401-2407.
18. Yaspan BL, Williams DF, Holz FG, et al. Targeting factor D of the alternative complement pathway reduces geographic atrophy progression secondary to age-related macular degeneration. Sci Transl Med 2017;9.
19. Khanifar AA, Lederer DE, Ghodasra JH, et al. Comparison of color fundus photographs and fundus autofluorescence images in measuring geographic atrophy area. Retina 2012;32:1884-91.
20. Duic C, Vance E, Agron E, Keenan TDL, Group AR. Alcohol consumption and risk of age-related macular degeneration and geographic atrophy progression: AREDS2 report 34. Ophthalmol Retina 2024;9:3:200-211.