Glycosylated hemoglobin and its temporal relationship to plasma glucose in non-insulin dependent (type 2) diabetes mellitus.

Total glycosylated hemoglobin (HbAI) and random plasma glucose were measured at monthly intervals for 6 mo in 33 non-insulin dependent (type 2) diabetics. The mean HbAI and mean plasma glucose in individual patients over the 6 mo showed a close correlation (p less than 0.001). A significantly higher HbAI (p less than 0.001) for a given plasma glucose was seen in those patients receiving combined chlorpropamide and metformin therapy (n = 14) than in those receiving chlorpropamide alone (n = 19). For each patient correlation coefficients were calculated between plasma glucose and HbAI with time lags of 0.1 and 2 mo. The coefficients with no time lag showed a significant tendency to be positive (p less than 0.01) whereas those with time lags of 1 or 2 mo were not significant. A constant proportional variability of both HbAI and plasma glucose over time was demonstrated, the mean coefficient of variation for HbAI being 8.4 +/- 2.7% and for plasma glucose 22.3 +/- 9.7%. We conclude that HbAI provides an index of plasma glucose control, not during the preceding few months as was previously thought but, during the previous few weeks. In terms of variability from month to month a single HbAI determination was equivalent to approximately the mean of 3 single plasma glucose values.