Abstract
Purpose: :
To describe 15–year cumulative incidence of macular holes and associated risk factors in a population–based study.
Methods: :
Of 4,926 persons, 43–86 years of age at the baseline examination in 1988–1990, living in Beaver Dam, WI, 3,684, 2,764, and 2,119 participated in 5, 10, and 15–year follow–up examinations, respectively. The presence of macular holes was determined by grading of stereoscopic color fundus photographs. Partial thickness holes were excluded from analysis. A standardized protocol including height, weight, blood pressure (BP), intraocular pressure (IOP), blood studies and medical history was performed at each examination. Cumulative incidence was estimated by the product–limit method, and multivariable analyses performed by linear logistic modeling.
Results: :
Of 3772 people at risk of developing a macular hole, there were 17 incident cases (15 monocular, 2 binocular). The 15–year incidence of macular holes was 0.6%. Final visual acuity decreased 29 logMar letters from baseline in eyes with holes compared to 5 letters in eyes without (p<0.001). Incidence of macular holes was 0.4%, 0.9%, 0.6%, and 0% in subjects 43–54, 55–64, 65–74, and 75–86 years of age, respectively (test for trend, p=0.55). There was no difference between women and men (p=0.70). In other univariable analyses, incidence of macular holes was higher in subjects with diabetes (1.6% vs 0.5%, p=0.05), history of cancer (1.6% vs 0.5%, p=0.003), higher body mass (0.9% in quartile 4 vs 0.3% in quartiles 1–3, p=0.03), lower diastolic BP (1.1% for <70mmHg vs 0.5% for ≥70mmHg, p=0.03), and larger retinal arterioles (1.4% in quartile 4 vs 0.4% in quartiles 1–3, p=0.001). Suggestive associations (0.05<p<0.10) occurred with lower total cholesterol, higher IOP, and larger retinal venules. No associations were found with systolic BP, pulse pressure, hypertension, HDL cholesterol, blood studies, cardiovascular disease history, smoking, alcohol consumption, epiretinal membranes, or lens extraction. In a multivariable analysis, incidence of macular holes was associated with larger retinal arterioles (odds ratio (OR) for quartile 4 vs 1–3, 4.64, 95% confidence interval (CI) 1.76, 12.19), a history of cancer (OR 4.31, 95% CI 1.51, 12.27), and larger body mass (OR for 1 kg/m2 1.09, 95% CI 1.02, 1.17).
Conclusions: :
The cumulative incidence of macular holes over the 15–year period while low, 0.6%, was associated with profound loss of vision. Few systemic factors were found to be associated with macular holes. The low incidence suggests the need for pooling data from large population–based studies to examine its epidemiology.