The purpose of the study was to estimate the 14-year incidence of visual loss in a diabetic population and to examine its relationship to potential risk factors.

Cohort study.

A population-based sample of younger onset diabetic persons diagnosed younger than 30 years of age and taking insulin (n = 880) were examined at baseline, 4 years, 10 years, and 14 years.

Visual acuity (VA) as measured by the Early Treatment Diabetic Retinopathy Study protocol was performed.

Visual impairment (VI), defined as a VA of 20/40 or worse in the better eye; blindness, defined as a VA of 20/200 or worse in the better eye; and doubling of the visual angle were measured.

Cumulative 14-year incidences of VI, doubling of the visual angle, and blindness were 12.7%, 14.2%, and 2.4%, respectively. In univariate analyses, loss of vision as measured by doubling of the visual angle is associated with older age, longer duration of diabetes, higher glycosylated hemoglobin, higher systolic and diastolic blood pressure, presence of proteinuria, more pack-years smoked, presence of macular edema, and more severe retinopathy. In logistic regression analyses, incidence of doubling of the visual angle is associated independently with retinopathy (odds ratio [OR], 1.07; 95% confidence interval [CI], 1.03, 1.11 for each level), glycosylated hemoglobin (OR, 1.46; 95% CI, 1.28, 1.66 for each 1%), proteinuria (OR, 2.32; 95% CI, 1.39, 3.88 for presence), and age (OR, 1.45; 95% CI, 1.20, 1.75 for 10 years). In addition, a change in glycosylated hemoglobin from baseline to the 4-year examination is associated with loss of vision (OR, 1.15; 95% CI, 1.02, 1.30 for a 1% increase).

Loss of vision continues to be significant in persons with diabetes. These results suggest that prevention of retinopathy through control of glycemia will have a beneficial effect on visual outcome.