Abstract
AIMS: In the Diabetes Control and Complications Trial, mean updated HbA(1c) accounted for most of the differential risk of microvascular complications between intensive and conventional insulin therapy. We hypothesized, however, that a more precise measure of chronic hyperglycaemic exposure may be the incremental area-under-the-HbA(1c)-curve above the Diabetes Control and Complications Trial-standardized normal range for HbA(1c) (iAUC(HbA1c>norm)). METHODS: Using the Principal Diabetes Control and Complications Trial data set, we compared the following three measures of chronic glycaemic exposure for their capacity to predict retinopathy, nephropathy and neuropathy during the Diabetes Control and Complications Trial: mean updated HbA(1c), iAUC(HbA1c>norm), and total area-under-the-HbA(1c)-curve (tAUC(HbA1c)). For each outcome, models using each of these three glycaemic measures were compared in the following three ways: hazard or odds ratio, χ(2) statistic, and Akaike information criterion. RESULTS: The three glycaemic measures did not differ in their prediction of neuropathy. iAUC(HbA1c>norm) was modestly superior to mean updated HbA(1c) for predicting nephropathy (χ(2) P = 0.017, Akaike P = 0.032). In contrast, for predicting retinopathy, both iAUC(HbA1c>norm) (χ(2) P = 0.0005, Akaike P = 0.0005) and tAUC(HbA1c) (χ(2) P = 0.004, Akaike P = 0.004) were significantly better than mean updated HbA(1c). Varying its HbA(1c) threshold incrementally between 37 and 53 mmol/mol (5.5-7.0%), inclusive, did not improve the prediction of retinopathy by iAUC(HbA1c>threshold) beyond that of tAUC(HbA1c,) consistent with the concept of a continuous relationship between glycaemia and retinopathy, with no glycaemic threshold. CONCLUSIONS: Both iAUC(HbA1c>norm) and tAUC(HbA1c) were superior to mean updated HbA(1c) for predicting retinopathy. Optimal assessment of chronic glycaemic exposure as a determinant of retinopathic risk may require consideration of both the degree of hyperglycaemia and its duration.