Relationship of hyperglycemia to the long-term incidence and progression of diabetic retinopathy.

BACKGROUND: The object was to examine the relationship of hyperglycemia, as measured by glycosylated hemoglobin level, to the incidence and progression of diabetic retinopathy over a 10-year period.
METHODS: Patients who were younger (n = 682) and older (n = 834) than 30 years at onset of diabetes participated in baseline (1980-1982) and follow-up (1984-1986 and 1990-1992) examinations of a population-based cohort study. Glycosylated hemoglobin levels were measured by microcolumn. Retinopathy was determined from stereoscopic fundus photographs.
RESULTS: Persons with glycosylated hemoglobin levels in the highest quartile at baseline were more likely to have progression of retinopathy than persons with levels in the lowest quartile (younger-onset group: relative risk [RR], 2.9; 95% confidence interval [CI], 2.3 to 3.5; older-onset group taking insulin: RR, 2.1; 95% CI, 1.6 to 2.8; and older-onset group not taking insulin: RR, 4.3; 95% CI, 3.0 to 6.2) and were more likely to develop proliferative diabetic retinopathy (younger-onset group: RR, 7.1; 95% CI, 4.6 to 11.1; older-onset group taking insulin: RR, 3.1; 95% CI, 1.5 to 6.1; and older-onset group not taking insulin: RR, 13.8; 95% CI, 4.8 to 39.5). These relations were significant (P < .005) in all groups examined, even after controlling for other risk variables.
CONCLUSIONS: These data are compatible with the hypothesis that long-term control of hyperglycemia, as measured by glycosylated hemoglobin levels, is a significant risk factor for the long-term progression of diabetic retinopathy and that lower levels of glycosylated hemoglobin, even later in the course of diabetes, may modify the risk imposed by higher levels earlier in the course of disease in people with both younger- and older-onset diabetes.