Can [HbA.sub.1c] be converted to mean blood glucose?

Attempts to translate hemoglobin Ale into mean blood glucose via a mathematical formula are likely to introduce substantial error, according to results from an analysis of continuous glucose monitoring data in 47 children with type 1 diabetes.

A recent multinational trial identified a formula for converting [HbA.sub.1c] to mean blood glucose, based on continuous glucose monitoring (CGM) data from 643 diabetic and nondiabetic subjects. Pending the final results of that study, the American Diabetes Association, the European Association for the Study of Diabetes, the International Federation of Clinical Chemistry and Laboratory Medicine, and the International Diabetes Federation have called for laboratories to report that mean glucose value in addition to the [HbA.sub.1c] itself (Diabetes Care 2007;30:2399-400).

[ILLUSTRATION OMITTED]

However, new data from the Diabetes Research in Children Network (DirectNet) Study Group suggest that although [HbA.sub.1c] does clearly reflect mean glucose, there is substantial variability in individual mean glucose concentrations for a given [HbA.sub.1c] value. Moreover, the strong relationship between [HbA.sub.1c] and the risk for diabetic complications may reflect not only mean blood glucose but also a patient's propensity to glycosylate other structural proteins, the study group said (Diabetes Care 2008;31:381-5). Dr. H. Peter Chase of the University of Colorado, Denver, was principal investigator of the study group.

The 47 patients, aged 4-18 years, included 28 pump users and 19 who took multiple daily injections using glargine as their basal insulin. They wore continuous glucose monitors for a mean of 115 hours per week. The mean glucose over the previous 3 months was 18 mg/dL per 1.0% [HbA.sub.1c], with considerable variation in glucose concentrations for any given H[bA.sub.1c]. For example, for an [HbA.sub.1c] of 7.0%, mean glucose ranged from 138 to 189 mg/dL. This relationship was similar between the pump users and those taking injections, and at both 3-month and 6-month visits.

"For any given [A.sub.1c] level, mean sensor glucose levels differed by up to 50 mg/dL or more, making the conversion of [A.sub.1c] into mean glucose equivalents as suggested by a recent American Diabetes Association consensus statement tenuous at best," the authors said. Two measures of an individual patient's rate of glycation--a simple ratio of mean glucose to [HbA.sub.1c] and a derived "glycation index"--remained relatively constant over the 6 months of the study, suggesting that individuals glycate proteins at different rates and that this tendency to be a fast or slow "glycator" persists over time within an individual, they explained.

After controlling for mean glucose over the prior 3 months, [HbA.sub.1c] was not significantly associated with measures of glucose liability (amplitude of excursions, rate of change, or percentage of sensor readings above 250-300 mg/dL), thus refuting the conclusions from other studies that high glucose values, such as those often seen postprandially, contribute disproportionately to the [HbA.sub.1c] value. "We found no evidence to contradict the simple hypothesis that [A.sub.1c] directly reflects the integral of glucose level over time," they said.

Funding for the study was provided by the National Institutes of Health; the continuous glucose monitors and test strips were provided by Abbott Diabetes Care. BY MIRIAM E. TUCKER Senior Writer