Methods and Sources

In 2001, a consensus meeting was convened by the National Eye Institute at which many of the world's leading ophthalmic epidemiologists created standard case definitions for the eye conditions included in this report. Data was obtained from a review of the major epidemiological studies with the cooperation of their authors (see Table of Sources below).

The number of individuals with each disease and the total number at risk were provided in five-year age increments by race and sex for the adult population from each of the studies. These age, race and sex-specific prevalence rates were then combined using a meta-analysis technique for reducing the overall variance of the pooled rate. Appropriate logistic regression models were fit to the age, race and sex-specific pooled prevalence rates. For the 2012 edition of this report, these models were applied to the state level U.S. Census 2010 populations for each year of age to arrive at the number of individuals with disease by race and sex for each state. The state-level prevalence rates were then derived by dividing the number of individuals with disease in each state by the total population in each state. To estimate the prevalence of disease in the "other" race category, the age and sex-specific rates for Whites, Blacks and Hispanics were averaged, then logistic regression models were developed based on these averages.

Obtaining an actual count of the number of eye disease cases in America would be virtually impossible. While these estimates do not represent exact measurements, they provide the best available information on the scope of the most serious threats to good vision for American adults in the 21st century.

Table of Sources

Primary prevalence rate source for: B = Blacks, W = Whites, H = Hispanics
B/VI = blindness and vision impairment, R/E = refractive errors, AMD = age-related macular degeneration, D/R = diabetic retinopathy