Five-year incidence and disappearance of drusen and retinal pigment epithelial abnormalities. Waterman study.

PURPOSE To obtain 5-year longitudinal data on age-related macular degeneration (AMD) that might be useful for disease prognosis, public health planning, and clinical trial development. PATIENTS AND METHODS Baseline (1985) and 5-year follow-up (1990) fundus photographs of 483 watermen over 30 years of age who participated in a cohort study conducted on the eastern shore of Maryland were graded independently in a reliable, standardized fashion. Eyes in which AMD appeared or disappeared also were graded in a side-by-side fashion. RESULTS Development of definite choroidal neovascularization and/or disciform scarring occurred in one of 50 participants over 70 years of age, specifically one of 15 participants over 70 years of age with AMD-3 (defined as large or confluent drusen focal hyperpigmentation of the retinal pigment epithelium [RPE], and/or nongeographic atrophy of the RPE). Appearance of large drusen, focal hyperpigmentation, or AMD-3 was age related, occurring in 5%, 1%, and 7%, respectively, of participants aged 50 to 59 years; 17%, 3%, and 14%, respectively, of participants aged 60 to 69 years; and 17%, 9%, and 26%, respectively, of participants aged 70 years or more. Disappearance of large drusen, hyperpigmentation, or AMD-3 occurred in 16 (34%) of 47 participants, 11 (58%) of 19 participants, and 17 (28%) of 61 participants, respectively, who had each feature photographically present in 1985. Among the 47 eyes identified in which AMD-3 developed by independent gradings, 38 cases of AMD-3 (81%) were confirmed on side-by-side grading. Among the 16 eyes identified as having AMD-3 that disappeared, nine disappearances (56%) were confirmed. Borderline differences in appearance of pigment, drusen size, drusen location, or photographic quality may have accounted for disappearance in seven cases (44%). CONCLUSIONS Prospective studies on the nonneovascular features of AMD (including large drusen and abnormalities of the RPE) must account for the appearance and disappearance of these features and support the idea that side-by-side gradings can complement independent gradings identifying appearance or disappearance of features of AMD.