Dietary Intakes of Eicosapentaenoic Acid and Docosahexaenoic Acid and Risk of Age-Related Macular Degeneration.

To evaluate the associations between intakes of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) and the intermediate and advanced stages of age-related macular degeneration (AMD).

Prospective cohort study.

We followed 75 889 women from the Nurses' Health Study and 38 961 men from the Health Professionals Follow-Up Study who were at least 50 years old, from 1984 to 2012 and 1986 to 2010, respectively. Cohort participants are mostly white (≥95%).

We assessed dietary intake by a validated food frequency questionnaire (FFQ) at baseline and every 4 years. We calculated cumulative average intakes of EPA and DHA from FFQs and also computed predicted erythrocyte and plasma scores directly from food intake using regression models. Cox proportional hazards models were used to compute the associations with AMD outcomes.

We confirmed 1589 incident intermediate and 1356 advanced AMD cases (primarily neovascular AMD) with a visual acuity of 20/30 or worse, owing primarily to AMD, by medical record review.

For intermediate AMD, the pooled hazard ratio (HR) between the 2 cohorts for DHA comparing the extreme quintiles of intake was 0.78 (95% confidence interval [CI], 0.66-0.92; P trend, 0.008) and for EPA + DHA was 0.83 (95% CI, 0.71-0.98; P trend, 0.03). The pooled HR for fatty fish, comparing ≥5 servings per week to almost never, was 0.61 (95% CI, 0.46-0.81; P trend, <0.001). For advanced AMD, the pooled HR for DHA was 1.01 (95% CI, 0.84-1.21; P trend, 0.75) and for fatty fish was 0.80 (95% CI, 0.59-1.08; P trend, 0.11). Secondary analyses using predicted erythrocyte and plasma scores of EPA and DHA yielded slightly stronger inverse associations for intermediate AMD and similar results for advanced AMD.

Higher intakes of EPA and DHA may prevent or delay the occurrence of visually significant intermediate AMD. However, the totality of current evidence for EPA and DHA and advanced AMD is discordant, though there was no association with advanced AMD in the present study.