Abstract
A novel fluorometric method for measurement of glycated serum albumin has been automated on a Cobas Bio centrifugal analyser. The method has been standardised, optimised and adapted to enable analysis of untreated serum samples and at a consumable cost of 25p per sample (including correction for serum albumin levels), 50-60 samples can be assayed in 1 hour. Between-batch imprecision was found to range from a CV of 5.8% to 11.4%, with a borderline diabetic/non-diabetic CV of 9.6%. Analysis of samples from a diabetic (n=86) and a non-diabetic (n=75) population produced a mean diabetic value 2-fold higher than the mean nondiabetic value and the assay results were shown to correlate well with an aminophenylboronate affinity chromatography method for glycated albumin (r=0.842). Optimum conditions for sample and reagent storage have been established and the effects of several possible interferents upon analysis and on in vitro glycation of albumin as detected by the fluorometric method have been investigated. Various chemical modifications of serum albumin have been performed in order to try and elucidate the chemical characteristics of this assay, and studies are described which suggest that glycated albumin is measured specifically. Together with affinity chromatography techniques for measuring glycated albumin and glycated total protein, and the established ‘fructosamine’ test, the new method has been assessed in various clinical situations. Components of biological variation have been determined for all these assays and the clinical utility of the fluorometric assay has been compared for monitoring both acute and chronic changes in glycaemia as well as diabetic pregnancy. The sensitivity of the new method as a screening test for diabetes/impaired glucose tolerance has been investigated and the suitability of the assay to a routine clinical chemistry laboratory has been established.