Abstract
Purpose: :
To determine whether the Age-Related Eye Disease Study (AREDS) simplified severity scale predicts the fifteen year incidence of late age-related macular degeneration (AMD) in the Beaver Dam Eye Study (BDES).
Methods: :
Of 4,926 persons, 43-86 years of age at the time of a baseline examination in 1988-1990, living in Beaver Dam, Wisconsin, 3,645 persons, with information collected in follow-up in 1993-95, and/or 1998-2000, and/or 2003-05 were included in this analysis. Stereoscopic 30° color fundus photographs of fields 1 and 2 were taken at each visit and graded in a masked fashion using the standardized Wisconsin Age-Related Maculopathy Grading Scheme. Two scores, one using the AREDS simplified severity scale based only on drusen size >125 um in diameter and/or presence of pigmentary abnormalities (PA) and the other the more detailed BDES-AMD grading that included drusen size and area, as well as presence of PA was assigned to each eye at baseline.
Results: :
Over 15 years there were 89 study participants who developed late AMD. Controlling for age, the cumulative incidence of late AMD accounting for the competing risk of death was 2.0% (95% CI 0.3 - 4.3%), 8.9 (3.1 - 14.6%), 16.0% (5.9, 26.1%), 34.7% (16.2 - 53.2%), and 32.3% (16.3- 48.4%) for individuals with a baseline AREDS clinical score of 0,1,2,3 and 4, respectively. In a similar analysis using the first 4 steps of the BDES severity scale the cumulative incidence was 1.0% (95% CI 0.5 - 1.5%), 5.2% (2.3 - 8.1%), 22.0% (13.4, 30.6%), and 39.5% (24.4 - 54.6%). The BDES 6-level AMD scale independently added to the predictive ability of the AREDS scale for late AMD (p<0.01) with an improvement of the R2 from 0.44 to 0.46..
Conclusions: :
The AREDS simplified AMD severity scale predicts progression to late AMD in the BDES cohort and while the BDES 6-step scale leads to a small improvement, both AMD severity scales result in similar findings regarding prediction of late AMD.
Keywords: age-related macular degeneration • clinical (human) or epidemiologic studies: prevalence/incidence