Age-Related Macular Degeneration and Mortality in the Age-Related Eye Disease Study (AREDS): The Effect of Sex and Time

Abstract

Purpose: Age-related macular degeneration (AMD) shares similar risk factors and pathogeneses with cardiovascular diseases (CVDs). Epidemiologic studies over the past 2 decades analyzing the association between AMD and all-cause and CVD-specific mortality have failed to yield conclusive results. The purpose of this analysis is to investigate the sex-specific association between AMD and all-cause and CVD-specific mortality, and to assess whether duration of follow-up alters the strength of association.

Design: The database of Genotypes and Phenotypes (dbGaP) data set for the Age-Related Eye Disease Study, a randomized clinical trial of high-dose antioxidants in AMD prevention, with participants enrolled from 1992 to 1998 and followed through 2005, was used in the analysis.

Participants: There were 4757 Age-Related Eye Disease Study participants aged 55 through 80 years (mean, 69.4 years; 44.1% male) recruited from 11 retinal specialty clinics. Participants had standard Age-Related Eye Disease Study AMD categories (category 1, n = 1117; category 2, n = 1062; category 3, n = 1621; category 4, n = 957).

Methods: The sex-specific adjusted hazard ratio (HR_adj) between baseline AMD and all-cause and CVD-specific mortality was determined at multiple time points (e.g., 5, 7, 10, and all years), adjusting for age, race, diabetes, hypertension, angina, cancer, smoking, obesity, clinical trial antioxidant treatment category, and education.

Main outcome measures: Sex-specific all-cause and CVD-specific mortality.

Results: Mean follow-up was 9.6 years (range, 0.5-12.5 years), with 1087 deaths (category 1, n = 197 [17.6%]; category 2, n = 200 [18.8%]; category 3, n = 356 [22.0%]; category 4, n = 326 [34.1%]). Sex-stratified models demonstrated sex differences; in women, a significant association between AMD category 4 and all-cause mortality existed compared with category 1 at each period (HR_adj, 1.5-2.3; all P ≤ 0.005); similar category 4 findings were present with CVD-specific mortality, strengthening with shorter periods (HR_adj, 1.9-4.6; all P ≤ 0.01). Among men, a significant association between all AMD stages and all-cause (HR_adj, 1.5-2.3; all P ≤ 0.05) and CVD-specific mortality (HR_adj, 1.6-4.0; all P ≤ 0.05) existed for nearly all periods.

Conclusions: Substantial late AMD cases and deaths exceed those in previous population-based studies to better test mortality-related hypotheses. Age-related macular degeneration was significantly associated with all-cause and CVD-specific mortality. Relationships weakened over a longer duration of follow-up, and sex seems to modify the association. Future analyses are warranted to interrogate the possible clinical usefulness of these relationships.