New spectrophotometric method for the determination of hemoglobin A1 compared with a microcolumn technique.

Abstract
We compared a spectrophotometric kit method (Glycospec) for determination of glycosylated hemoglobin (HbA1) with a microcolumn kit method (Bio-Rad). The Glycospec method is based on the change in absorbance when phytic acid binds to hemoglobin A. With glycosylated hemoglobin there is no such change because the binding is blocked by the sugar moiety. Inter-assay CVs were 2-6% for both methods. In healthy subjects the mean (+/- SD) value for HbAl was about 1% higher with the spectrophotometric than the microcolumn method. For samples from 122 diabetics the correlation between values for HbAl obtained with the two methods was acceptable (r = 0.89), although the spectrophotometric technique yielded 2-4% higher values, a difference at least partly due to the absence of 2,3-diphosphoglycerate from the spectrophotometric standards. Adding 1.8 mmol of it per liter to these standards caused displacement of the standard curve; HbAl values then agreed well with those of the microcolumn method. The spectrophotometric procedure is easily automated, and therefore is well suited for large-scale analyses if problems with standards and calibration can be solved.