Purchase this article with an account.
J.M. Rosenthal, J.P. SanGiovanni, E. Agron, G. Reed, E. Chew, J. Seddon, F. Ferris, AREDS Group; A Dietary Antioxidant Index and Risk for Advanced Age–Related Macular Degeneration in the Age–Related Eye Disease Study . Invest. Ophthalmol. Vis. Sci. 2006;47(13):1134.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Nutrients with antioxidant (AO) properties interact biochemically and biophysically in the retina. It is important to consider the collective contribution of such nutrients in risk estimation for retinal diseases potentially associated with oxidative stress.
To examine the relationship of a dietary AO index with neovascular (NV) AMD.
We used a nested case–control design to compare a derived AO index in AREDS subjects with baseline NV AMD (n=658) to subjects without AMD (n=1115). NV AMD was determined at a reading center from stereo fundus photographs. Standardized questionnaires yielded demographic, lifestyle, medical, and ocular data. We estimated baseline dietary intake with a validated food frequency questionnaire and selected elements of our index from a set of nutrients with AO properties that were also associated with NV AMD in multivariable–single nutrient models containing known covariates. Nutrients included lutein/zeaxanthin (L/Z), b–cryptoxanthin, b–carotene, vitamins C and E, selenium, and zinc. We applied principal components analysis to this nutrient set and computed the AO index score for an individual by multiplying the standardized intake of a nutrient by its factor loading score and summing across all nutrients. The lowest quintile of the AO index was the referent group for other quintiles.
Our first principal component accounted for 52% of the total variance in AO intake. L/Z contributed 13% to the total variance in the AO index; the other nutrients each contributed 12%–16%. In multivariable logistic regression models containing demographic, lifestyle, medical, and ocular factors associated with NV AMD, ORs (95%CIs) for AO index quintiles 5, 4, 3, and 2 versus quintile 1 were 0.5 (0.3–0.8), 0.5 (0.3–0.8), 0.6 (0.4–0.9), 0.8 (0.6–1.2), respectively. The likelihood of NV AMD was reduced by 16% to 40% in the single–AO nutrient analyses among people reporting highest levels of intake. The strongest association was for L/Z (OR=0.6; 95%CI 0.4–0.8).
People with highest values on a dietary AO index derived from principal components analysis were less likely to have NV AMD than their peers with lowest AO index values after considering the influence of known covariates. As compared with analyses on individual nutrients, this AO index (representing approximately equal contributions from 7 nutrients) yielded lower ORs and may have led to small gains in precision of our estimates. Results might be affected by uncontrolled confounding and misclassification of exposure.
This PDF is available to Subscribers Only