Four assay methods for glycated hemoglobin compared as screening tests for diabetes mellitus: the Islington Diabetes Survey.

Abstract

We assessed the utility of four methods of glycated hemoglobin assay (agar gel electrophoresis with Schiff base, agar gel electrophoresis with prior removal of Schiff base, boronate affinity chromatography, and isoelectric focusing) as screening tests for diabetes mellitus, studying 223 subjects undergoing an oral glucose-tolerance test after fasting and 2 h after ingestion of 75 g of glucose. Assessment of glucose tolerance status according to the 1980 World Health Organization diagnostic criteria indicated that 13 subjects (5.8%) had diabetes, 48 (21.5%) had impaired glucose tolerance, and nine (4.0%) could not be classified (six of these because of missing values). Use of receiver-operating characteristic curves to compare the assays as screening tests showed the affinity chromatography assay to be superior. Assays for glycated hemoglobin after fasting had better precision than post-glucose-load assays as screening tests, and test characteristics of glycated hemoglobin assays on fasting subjects were similar to those of the blood-glucose estimation after fasting. We conclude that measurement of glycated hemoglobin may be a useful screening test for diabetes, and that such measurement of stable glycated hemoglobin is as accurate as measurement of fasting blood-glucose in screening for diabetes. For this use, we found methods that measure stable glycated hemoglobin superior to that measuring both stable and labile Schiff-base fractions.

The relative merits of haemoglobin A1c and fasting plasma glucose as first‐line diagnostic tests for diabetes mellitus in non‐pregnant subjects

“…10 There are, however, considerable differences between the levels of diagnostic sensitivity and specificity reported for glycated haemoglobin in the literature. In our study, and in others, [11][12][13] HbA 1c was not as sensitive at high specificity as has been claimed elsewhere. 14 These differences may, in part, be attributable to glycated haemoglobin method differences, but most of the available evidence suggests that glycated haemoglobin is slightly less sensitive than FPG for diagnostic purposes.…”

Unexplained variability of glycated haemoglobin in non-diabetic subjects not related to glycaemia

“…We found that glycated haemoglobin assay imprecision and the variability of the glucose tolerance test [31][32][33] were not the major determinants of the poor relationship. Glycated haemoglobin levels measured by all assay methods showed a considerable range in our study sample [14], probably as a result of the heterogeneity of the population, for example in terms of age, ethnic origin and social class, and thus high and low glycators may demonstrate differences of levels of glycated haemoglobin of 1-2% at the same degree of glucose intolerance (Fig. 1).…”

“…Our own observations confirm the previous finding [28] that levels of glycated haemoglobin are stable over this period of time. The withinand between-assay coefficients of variation of these assays on ten replicate samples have been reported [14]. Haemoglobin levels were measured colorimetrically on an EDTA sample using a Coulter haemoglobinometer (Coulter Electronics Ltd., Luton, Bedfordshire, UK) [29].…”

“…In the Islington Diabetes Survey [10][11][12] a sample of 223 subjects had two glucose tolerance tests, and estimates of glycated haemoglobin on two occasions, on the second of which the assay was performed by four different methods on both the fasting and 2 h blood samples [12][13][14]. Wide differences in levels of glycated haemoglobin relative to the degree of glucose intolerance were found in this population, but levels of glycated haemoglobin were more closely interrelated, even when analysed by methods relying on different physicochemical properties.…”

A labile form of hemoglobin A1c is higher in African‐American youth with type 1 diabetes compared to Caucasian patients at similar glucose levels

“…15,16 This labile fraction was a nuisance as it reflected short term concentrations in glucose and its presence could confound the interpretation of HbA1c as an estimate of long-term MBG. [15][16][17][18] Taking advantage of its unstable nature, methods to remove this labile fraction such as pre-incubating samples in low pH and/or glucose-free buffers were introduced so as to measure the stable form of HbA1c alone. 15,16,[18][19][20] Other than its role confounding measurement of HbA1c, identifying and interpretation of labile HbA1c has been of passing clinical interest.…”