This article is more than 5 years old. Information may no longer be current.
Should we assess measures of glycemia other than HbA1c in pediatric patients?
Click Here to Manage Email Alerts
Although HbA1c has been the gold standard for assessing glycemic control in diabetes in the past several decades, concerns remain regarding its standardization and the accuracy of different methodologies for its determination.
The National Glycohemoglobin Standardization Project certifies the accuracy of laboratory methods for determining HbA1c based on Diabetes Control and Complications Trial standards. During the past 20 years, the accuracy of certified methods has greatly improved. There are, however, significant biases between point-of-care measurements and simultaneous clinical laboratory measurements. Bias increases with increasing HbA1c values.
|
|
Kamps and colleagues have reported that HbA1c values based on mean blood glucose readings (30-day meter downloads) can be different from measured HbA1c levels in some children. This difference is termed the hemoglobin glycation index (HGI), and it has been shown that black children may have significantly higher HGI than white children, suggesting biologic variability in HbA1c determinations.
Cohen and colleagues have coined the term “glycosylation gap” to describe the difference between laboratory-measured HbA1c levels and those predicted from fructosamine determinations. HGI and glycosylation gap are highly correlated and can be used to characterize individuals into high, moderate or low glycater groups. Estimated average glucose levels, published from the A1c-Derived Average Glucose Study Group are now recommended by the American Diabetes Association as a surrogate for HbA1c reporting. However, the estimated average glucose underestimates mean blood glucose in low HGI patients and overestimates mean blood glucose in high HGI patients.
Determination of serum fructosamine is useful in individuals with hemoglobinopathies associated with rapid red blood cell turnover. Fructosamine reflects average blood glucose levels during a 2- to 3-week period. Anhydroglucitol levels reflect glycemia over a few days and are negatively correlated with average blood glucose levels. However, recent unpublished data in children wearing continuous glucose sensors demonstrate that neither is more significantly correlated with mean blood glucose than HbA1c.
HbA1c levels do not reflect glycemic variability, and glycemic variability may contribute to both short- and long-term complications of diabetes. Glycemic variability can be determined from self-monitoring blood glucose data and from continuous glucose sensor data. A variety of indexes, including standard deviation, have been suggested as markers of variability. However, two of the most frequently used indexes — standard deviation and mean amplitude of glycemic excursion — are insensitive to hypoglycemia, whereas M-value and lability index are insensitive to hyperglycemia. To date, only the average daily relative risk predicts both glycemic extremes.
Statistical analysis of continuous sensor data must take into consideration the dependent nature of the data points produced. Deterministic methods should be used to describe these time-series data. Methods for assessing variability from these data, which take these properties into account, include rate of change, continuous overall net glycemic action, risk analysis, Poincaré plots and variability grid analysis.
Practitioners without access to readily available alternative determinations of glycemia or sophisticated data management systems will most likely continue to rely on point-of-care HbA1c values to monitor their patients’ glucose control. A relatively simple computation of percent high and percent low blood glucose (or percent within a set target blood glucose zone) from logbook entries may be a useful adjunct to assess and track glycemic variability in their patients.
William L. Clarke, MD, is Robert M. Blizzard professor of pediatrics at University of Virginia Health System.
For more information:
- Clarke W. Diabetes Technol Ther. 2009;11:S45-S54.
- Cohen R. Diabetes Care. 2003;26:163-167.
- Hanas R. Pediatr Diabetes. 2010;11:209-211.
- Hempe J. Diabetes Care. 2010;33:1449-1451.
- Holmes E. Am J Clin Pathol. 2008;129:540-547.
- Kamps J. Diabetes Care. 2010;33:1025-1027.
- Kovatchev B. Diabetes Care. 2006:29:2433-2438.
- Little R. Curr Opin Endocrinol Diabetes Obes. 2009;16:113-118.
- Nathan D. Diabetes Care. 2008;31:1473-1478.
- Nguyen T. Pediatr Diabetes. 2007:8:214-219.