Is fasting glucose variability a risk factor for retinopathy in people with type 2 diabetes?

AIMS: Fasting plasma glucose variability strongly predicts the incidence of cardiovascular events in type 2 diabetic patients. We prospectively assessed whether fasting plasma glucose variability predicts the development/progression of retinopathy in a large cohort of type 2 diabetic outpatients. METHODS: In the period 1996-1999, 1019 type 2 diabetic participants (aged 69+/-11 years) in the Verona Diabetes Study underwent at least 3 fasting plasma glucose (FPG) determinations and an eye examination by retinography. Of these, 746 underwent a 2nd eye examination in the period 2000-2004, while 273 did not (102 patients had died before undergoing the 2nd eye examination). For each patient, the mean (M-FPG) and the coefficient of variation of FPG (CV-FPG) were computed. RESULTS: By the 2nd eye examination, 124 patients had either developed new retinopathy (79 patients) or progressed to a more severe degree of retinopathy (45 patients). In a multivariable logistic regression analysis, the development/progression of retinopathy was independently predicted by average glycaemia over time, expressed as glycated haemoglobin (odds ratio [OR] 1.82, 95%CI 1.40-2.38 for 1 SD increase) or M-FPG (OR 1.88, 1.47-2.41), but not by CV-FPG. Among other independent variables, HDL-cholesterol was inversely associated with the development/progression of retinopathy. CONCLUSIONS: These results suggest that in elderly type 2 diabetic patients the magnitude of hyperglycaemia, but not fasting plasma glucose variability, strongly predicts the development/progression of diabetic retinopathy independently of other known risk factors.