Haemoglobin A1c--a marker for complications of type 2 diabetes: the experience from the UK Prospective Diabetes Study (UKPDS).

Haemoglobin A1c (HbA1c) is the pre-eminent factor for quantifying the risk of complications in patients with diabetes and for monitoring glycaemia. Intervention to lower blood glucose in the two landmark clinical trials, the UK Prospective Diabetes Study (UKPDS) and the Diabetes Control and Complications Trial (DCCT), led to a reduction in the microvascular complications of diabetes. Glycaemic status could be compared in the UKPDS and DCCT as the Bio-Rad Diamat HPLC analyser, as used in the DCCT, was introduced in 1989 for measurement of HbA1c in the UKPDS, after liaison with the DCCT. Results from other methods used for measurement of glycated haemoglobin during the UKPDS were aligned to this method. The Bio-Rad Diamat analyser in the central laboratory for the UKPDS, reference range 4.5-6.2% HbA1c, was certified as comparable to the DCCT by the National Glycohemoglobin Standardization Program in 1998. A median difference in HbA1c of 0.9% was maintained over 10 years between the intensively and conventionally treated groups in the UKPDS (7.0% vs. 7.9% HbA1c) despite HbA1c increasing over time. Clinical care was transferred to general practitioners after the end of the main glucose control study for post-study monitoring. Over the first 3 years of post-study monitoring, HbA1c rose slightly in the previously intensively treated group with no appreciable increase in the conventional group, due to intensification of therapy. At near-normal HbA1c, < 6%, the updated mean value chosen to reflect glycaemic exposure throughout the UKPDS, the incidence of myocardial infarction was 2-3 times that of microvascular disease, with similar incidences for both complications at >10% updated mean HbA1c. Relationships between the risk of complications of type 2 diabetes and updated mean HbA1c had no observable thresholds. The UKPDS risk engine derived from the UKPDS database calculates coronary heart disease risk using HbA1c as a continuous variable and could now replace the Framingham equations for patients with type 2 diabetes.