Longitudinal studies of incidence and progression of diabetic retinopathy assessed by retinal photography in pima indians.

OBJECTIVE To examine incidence and progression of retinopathy using retinal photographs in Pima Indians and to compare the results with those obtained when retinopathy is assessed by direct ophthalmoscopy. RESEARCH DESIGN AND METHODS We undertook an analysis of examinations conducted between 1 April 1982 and 31 December 1990 in residents of the Gila River Community in central Arizona. Data were taken from 411 people with diabetes who had at least two examinations during this period. Incidence and progression of retinopathy were evaluated by retinal photography and fundoscopy, and hazard rate ratios were calculated for various potential risk factors measured at baseline. RESULTS Previously diagnosed retinopathy tended to progress, except in individuals with minimal nonproliferative retinopathy, among whom follow-up examinations were more likely to show no retinopathy. Diabetes duration (hazard rate ratio [HRR], 1.06 per year difference [P = 0.007]), hyperglycemia (HbA(1) HRR, 1.27 per 1% difference [P < 0.0001]), the type of treatment for diabetes (insulin use HRR, 3.06 [P = 0.0007], and oral hypoglycemic use HRR, 2.40 [P = 0.0034], compared with individuals on no pharmacotherapy), and macroalbuminuria (HRR, 2.86, compared with individuals without macroalbuminuria [P = 0.0486]) were associated with the development of retinopathy. Although fundoscopy detected fewer cases of retinopathy, HRRs for most risk factors were similar when retinopathy was assessed by fundoscopy rather than retinal photography. CONCLUSIONS Although retinopathy tends to worsen over time, some eyes show improvement, especially in patients with minimal nonproliferative retinopathy. As in other populations, glycemic control is the major modifiable risk factor for the development and progression of retinopathy.