The AREDS formula, which contains ~5, ~6 and ~18 times the Recommended Dietary Allowance (RDA) levels of zinc, vitamin C and vitamin E, respectively, was shown to be protective against progression to advanced AMD in the AREDS cohort, in which a high proportion (67%) of the participants took an additional RDA-level of these nutrients from Centrum during the trial period.4
Given the high levels of the antioxidants from the supplements, it is not surprising that we found that no additional benefit accrued from intake of these nutrients from foods to that offered by taking the AREDS supplement. This finding suggests that there may be a saturating level for these nutrients with respect to macular health.
Because of the potential hazard of lung cancer from beta-carotene, 20
the AREDS supplement is not suggested for smokers. Our observation that high levels of dietary beta-carotene intake may increase the risk for progression to advanced AMD further supports the deletion of beta-carotene from the AREDS formula ( and Results). Because half of the AREDS AMD trial participants received a high dose of beta-carotene from the intervention, and 67% of them took Centrum, the average total intake level for beta-carotene in this cohort was much higher than in other populations. Caution must be taken in interpreting our findings because they were of marginal statistical significance and may be due to the large number of analyses that have been done. However, a recent study from the Blue Mountains Eye Study, Australia, published during the review period of the present study, found that beta-carotene intake from diet alone predicted neovascular AMD, and the association was evident in both ever and never smokers.7
Our results showed a protective effect against progression to advanced AMD in those with the highest (>75th percentile) consumption of omega-3 fatty acids (DHA and EPA) (). This may reflect the fact that Western diets provide low levels of omega-3 fatty acids, and suggest that increasing levels of omega-3 fatty acid intakes would enhance prevention of advanced AMD in a population consuming a Western diet. In our cohort, the 75th percentiles of DHA and EPA intakes were 64.0 and 42.3 mg/day, respectively, which are much lower than the recommended total intake level (650 mg/day) for DHA plus EPA. A dietary strategy for achieving this level is to consume two to three servings per week of cold-water fatty fish, such as salmon, tuna, mackerel, sardines, shellfish and herring.21
Like other western-diet populations, our cohort consumed a large amount of refined carbohydrates (high-GI foods). These data show that a 6-unit reduction in dGI, which is approximately equivalent to replacing the five slices of white bread (GI = 100) with five slices of whole-grain bread (GI = 79) from the daily diet for a person consuming 250 g of total available carbohydrate per day, might eliminate almost 8% of advanced AMD cases in a 5-year period.12 13
Interestingly, the antagonistic DHA-AREDS supplement interaction () was not observed in our analysis of progression to advanced AMD (). These results may imply that the protective effect of DHA may be through different mechanisms at different stages of AMD progression. Furthermore, taking together the results from early () and advanced () AMD analyses, these data also suggest that the timing of different supplementations may be important. One may speculate that a diet rich in omega-3 fatty acids or omega-3 fatty acid supplementation, without the AREDS supplement, might be most protective against progression of early AMD if started at the earliest stages of AMD, while, when combined with the AREDS supplement, it might be most protective with regard to preventing progression to advanced AMD.
It is hypothesised that high-GI diets induce higher post-prandial glycoxidative stress through the formation of advanced glycation end products, glycoxidation, accumulation and precipitation of glycated protein aggregates, and subsequent inflammatory and angiogenic responses.11-13 22 23
In addition, because compensatory hyperlipidaemia in the late postprandial stage after consuming high-GI foods may also play a role in the pathogenesis of AMD, and the protective effect of omega-3 fatty acids may be through modulating the postprandial blood lipid profile,11-13 24
the synergistic interactions of low-GI diets with AREDS supplement () and high omega-3 fatty acid diet () provide further support to our hypothesis.
There may be a concern about dietary change over the study period. However, information derived from FFQs has been considered as valid for measuring long-term dietary intake in epidemiological studies,18
and the non-differential misclassification of dietary intake would modify our result toward no association. Residual confounders could be another concern; however, as discussed in our previous paper,13
it is unlikely that they can totally explain our findings.
The present study adds additional support to the idea that diet and, if necessary, supplementation can be optimised for the prevention of AMD. However, some issues require further study. First, because adverse effects may appear only after long-term use, and the effect of an intervention may vary by different stages of disease development, the duration and timing of an intervention should be evaluated carefully. Nutrient–nutrient interactions may also influence the effect of an intervention at different stages of disease progression and vary in different populations. Although new trials will offer the soundest answers, observational data from post-trial follow-up studies may offer valuable information to address these issues and to plan the next generation of trials.