The Importance of Determining Irreversibly Glycosylated Hemoglobin in Diabetics

Abstract

The concentration of glycohemoglobins (HbA1(a+b+c), HbA1,) was measured before and after incubation of normal and diabetic erythrocytes for 6 h at 37°C in saline. This procedure removes as much as 80-90% of the labile glucose-HbA0 adduct (labile HbA1,), thus allowing accurate estimation of irreversibly glycosylated hemoglobin (stable HbA1,). The concentration of HbA1 measured before such an incubation is total HbA1, (stable + labile). We determined the concentration of total, stable, and labile HbA1, in the same blood samples used to measure fasting plasma glucose (FPG) every day, for 4 consecutive days, in two groups of hospitalized insulin-treated diabetics. Group A subjects (N = 7) were type I, C-peptide negative, unstable diabetics, while group B subjects (N = 15) were type II, C-peptide positive, stable diabetics. Individual day-to-day variations of total HbA1, were wide in group A (Δ = 1.58 ± 0.14%), and slight in group B (Δ = 0.12 ± 0.01%; P < 0.001), paralleling similar plasma glucose fluctuations. Day-to-day variations of stable HbA1, were virtually absent not only in group B subjects with stable glycemic values (Δ = 0.08 ± 0.01%), but also in those of group A with marked glycemic instability (Δ = 0.07 ± 0.01%; P = NS). Day-to-day variations of labile HbA1, were marked in group A (Δ = 1.31 ± 0.14%), but negligible in group B (Δ = 0.15 ± 0.03%; P < 0.001). On admission, FPG correlated with labile HbA1, in group A (r = 0.89) and B (r = 0.71). FPG correlated with stable HbA1, in group B (r = 0.73) but not in group A subjects and with total HbA1, more closely in group B (r = 0.73) than in A (r = 0.61). A very close correlation was found between the concentration of total and labile HbA1 in subjects of group B (r = 0.82). In group B, fasting, post-breakfast, and mean daily plasma glucose values, determined every 3-6 days during the 2 mo before admission, significantly correlated both with total and stable HbA1 determined on admission, while in group A they did not. In group A, the correlation was significant when stable instead of total HbA1 was considered. We conclude that the significant fluctuations of total HbA1 reduce its value as an index of long-term control in unstable diabetics. On the other hand, a single determination of stable HbA1, totally independent of simultaneous blood glucose values, closely reflects blood glucose control over the previous 2 mo. We propose routine estimation of stable HbA1, which is simple and straightforward, to carry out follow-up studies of unstable diabetics.