Abstract
A method is presented for the estimation of fructosamine using a Cobas Bio centrifugal analyser. The effect of three different preparations of human serum albumin used for construction of calibration curves of 1-deoxy-1-morpholinofructose is described. The selection of serum albumin and the concentration used in the standard solutions is critical since the dose-response curve is affected differently and will therefore influence the estimated values. Normal ranges were obtained for non-diabetic subjects with normal protein status and for a group of females with reduced albumin levels due to pregnancy or oestrogen therapy. There was no significant difference between fructosamine levels in these populations. Fructosamine was also estimated in 250 patients attending a diabetic out-patient department and this correlated well with haemoglobin A1 estimated simultaneously. The method is rapid, technically simple and inexpensive and may prove to be a useful and reliable alternative to HbA1 estimation.