eye risk before diagnosis

A recent study proposes new thresholds for predicting risk of diabetic retinopathy. Using more precise gradation in the "good control" or "prediabetic" range than comparable earlier research, Massin et al. found the following correlations among 700 French volunteers studied for ten years, and were confident in proposing them as predictive:

HbA1c %	fasting plasma glucose mg/dL	predicted retinopathy
6.0	—	6.0%
—	108	8.4%
—	116	14.0%
6.5	—	14.8%

The takeaway? A difference in the range of half a percentage point of HbA1c, or 8 mg/dL of fasting plasma glucose, roughly doubles one's risk of potentially blinding retinopathy, from about one in 15 to one in 7. This risk correlation held even among subjects who did not develop diabetes, as currently defined, over the course of the study, retaining their diagnosis of "impaired fasting glucose".
The current diagnostic threshold for diabetes is an HbA1c of 6.5, the highest value in the proposed predictive model. Since the diagnostic threshold in terms of fasting blood glucose is 126 mg/dL, the remaining values are well below the threshold, in the so-called "prediabetic" range. Normoglycemic, neither diabetic nor prediabetic, is currently defined (at the upper end) as under 100 mg/dL fasting glucose or under 5.7% HbA1c.
I have two reasons for highlighting this study:

• Absent a risk of going hypoglycemic—into dangerously low blood glucose levels—there is a compelling argument for driving blood glucose levels down even if you are being told, as people in the "prediabetic" range often are, that they have "good control". A level of control that may double my risk of blindness doesn't seem to me to have much good about it. (Type 1 diabetics or others considered to be at higher risk for "going hypo" may want to check with their care team about the extent to which tight control might mitigate both short- and long-term risks.)
• If diabetes is a progressive, degenerative, debilitating, and potentially fatal disease with multiple comorbidities—it is—and if "prediabetes" is an independent mortality risk whose defining numbers successfully predict so-called "complications", or more precisely late-stage symptoms, of diabetes—it is and they do—why on Earth is the diabetic establishment still sugar-coating the reality of this disease by mucking around with terms like "prediabetes"? When will medicine catch up with science and lower the diagnostic threshold? It has done so before, and can do so again. Come the day!

source: Massin P et al. Archives of Ophthalmology, 2011: 129(2): 188-195