Interval between onset of mild nonproliferative and proliferative retinopathy in type I diabetes.

Objective: To describe the interval between first appearance of mild nonproliferative diabetic retinopathy (NPDR) and first appearance of neovascularization (NV) in type I diabetes.

Setting: A longitudinal study of 269 patients followed up annually.

Participants: Participants had insulin-dependent diabetes and were free of proliferative diabetic retinopathy in both eyes at the baseline visit.

Main Outcome Measure: Stereoscopic color fundus photographs of each eye at each study visit, graded for features of retinopathy.

Results: Among the 305 eyes for which the duration of diabetes at the first appearance of mild NPDR could be determined, NV developed in 28 by the end of the study. Survival analysis showed that the later the onset of mild NPDR was, the shorter the time from onset of mild NPDR to onset of NV (relative hazard for each additional year to onset of mild NPDR, 1.22; 95% confidence interval, 1.10-1.35). Adjustment for systolic and diastolic blood pressure, proteinuria, and glycosylated hemoglobin (Hgb A10) levels did not change the relative hazard estimate for onset of mild NPDR. Higher levels of Hgb A10 were associated with a shorter time from onset of mild NPDR to onset of NV (relative hazard, 1.26; 95% confidence interval, 1.05-1.51 [after adjusting for time at onset of mild NPDR]), as were higher levels of diastolic blood pressure, although not significantly (relative hazard for 10-mm Hg increase in diastolic blood pressure, 1.52; 95% confidence interval, 0.82-2.83 [adjusting for onset of mild NPDR, Hgb A10 level, systolic blood pressure, and proteinuria]). Neither proteinuria nor systolic blood pressure had an effect on time from onset of mild NPDR to onset of NV, after adjustment for time at onset of mild NPDR, Hgb A10 level, and diastolic blood pressure.

Conclusion: Later onset of mild NPDR is not necessarily associated with delayed development of NV in patients with type I diabetes. Caution must therefore be used in assessing the value of interventions that delay the onset of mild NPDR without evidence of delayed onset of NV.