| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 University of Missouri School of Medicine, Columbia, MO 65212;
2
McNeil Specialty Products Company, New Brunswick, NJ 08903
aaddress correspondence to this author at: Department of Child Health, University of Missouri–Columbia, 1 Hospital Dr., M772, Columbia, MO 65212; fax 573-884-4748, e-mail RohlfingC{at}health.missouri.edu
Glycohemoglobin (GHb) is a measure of long-term mean glycemia that predicts risks for the development and/or progression of diabetic complications in patients with type 1 and type 2 diabetes (1)(2). Several reports have suggested, however, that although the within-subject variation in GHb unrelated to glycemia is minimal, there is substantial between-subject variation in GHb, e.g., "low glycators" and "high glycators" (3)(4)(5). These reports have suggested that because of this large between-subject variation, GHb may not be useful for diabetes screening or diagnosis and that when GHb is used for routine management of patients with diabetes, different patients may require very different GHb target values to achieve the same overall glycemic status. We therefore examined the biological variation of GHb and fasting plasma glucose (FPG) in nondiabetic individuals.
Individuals without diabetes (n = 48) participated in a study of an artificial sweetener that has no effect on GHb or plasma glucose concentrations [Submission to Food and Drug Administration. McNeil Specialty Products Company food additive petition 7A3987 (Sucralose), 1987–1997]. Because the study was designed to detect minimal changes in plasma glucose concentrations, all participants were men to avoid the effects of cyclic hormonal changes on insulin (and therefore, plasma glucose) concentrations. At the prestudy screening, all individuals were healthy on the basis of a medical history, physical examination, and electrocardiography results; results of hematology and blood chemistry studies, urine examination, and measures of blood glucose control (FPG, insulin, C-peptide, and hemoglobin A1c) were all within their respective reference intervals. Participants who failed a baseline oral glucose tolerance test [fasting >7.8 mmol/L (140 mg/dL), 1 h >11.1 mmol/L (200 mg/dL), and/or 2 h >7.8 mmol/L (140 mg/dL)] were excluded. Those who took medications that could affect glucose metabolism or who failed a drugs-of-abuse screen, had a history of a gastrointestinal disorder, or had a history of consuming more than two alcoholic drinks per day were also excluded. Serial samples for FPG and GHb analysis were collected by venipuncture after a minimum 8-h overnight fast on a weekly basis for a total of 12 visits. Three men with <10 data points for either FPG or GHb were excluded from the analysis. The study received approval from the institutional review boards of all participating study centers, and all individuals gave informed consent before their participation.
GHb values were measured by HPLC (Primus boronate affinity; interassay CV <3%) (6). FPG values were measured by a hexokinase assay (Roche Cobas Mira; interassay CV <3%) (7). SAS software was used to perform all statistical analyses; linear regression analysis examined the correlation between initial FPG and GHb values. We estimated the between-subject (Sg2), within-subject (Si2), and assay (Sa2) components of the total variance by a nested ANOVA (8) using the SAS Proc Varcomp software; restricted maximum likelihood was the method of estimation. We calculated the within-day component of assay variance (within-day Sa2) by combined within-run variance estimates for quality-control specimens analyzed two to five times in each analytic assay.
The mean, minimum, and maximum GHb values for each participant are shown in Fig. 1
. The correlation between initial GHb and FPG values was weak but statistically significant (r2 = 0.102; P <0.05). Table 1
shows the estimated variance components for GHb and FPG. The Sg2 component for GHb was much larger than the Si2 component. The mean GHb value was 4.9%; the between-subject SD (Sg) was 0.20% GHb (CVg = 4.0%). Thus, the between-subject mean ± 2 SD interval (95% confidence interval) was 4.5–5.3% GHb. The within-subject SD (Si) was 0.08% GHb (CVi = 1.7%), and the between-day assay SD (between-day Sa) was 0.11% GHb (CVa = 2.3%). The Si2 component of variation included the within-day Sa2 and Si2 because we were unable to directly separate the two components (the specimens were not analyzed in duplicate). However, the estimated within-day analytic SD (within-day Sa), based on quality-control data, was 0.07% GHb (within-day CVa = 1.5%), which indicates that most of the estimated CVi was attributable to the within-day Sa2 rather than Si2.
|
|
For FPG, Si2 and Sg2 were comparable. The mean FPG value was 5.3 mmol/L; the estimated Sg, Si, and between-day Sa SDs were 0.31 mmol/L (CVg = 5.8%), 0.30 mmol/L (CVi = 5.7%), and 0.09 mmol/L (between-day CVa = 1.7%), respectively. The estimated within-day Sa, based on quality-control data, was 0.04 mmol/L (within-day CVa = 0.8%), which indicates that most of the estimated CVi was attributable to Si2.
Our data show that for GHb, Si2 in nondiabetic individuals is minimal. Although we were unable to obtain a precise estimate of Si2 separate from Sa2, the CVi was likely <1%. Although the between-subject component was the largest component of the total variation in GHb, the between-subject mean ± 2 SD range in GHb for nondiabetic men was <1% GHb after accounting for Sa2.
A previous large-scale study has shown that GHb reliably categorizes glycemic control in nondiabetic individuals (9), and several studies have shown correlations between GHb concentrations and outcome risks in both diabetic (1)(2) and nondiabetic (10)(11) persons. Such findings suggest that the between-subject differences in GHb are mainly attributable to differences in mean glycemia rather than other factors.
Several factors may explain the relatively poor correlations observed between FPG and GHb values in nondiabetic individuals (3)(12). Both Si2 and Sg2 were higher for FPG than for GHb, and the intervals for these variables in nondiabetic individuals are relatively narrow. We also note that GHb is a more comprehensive measure of mean glycemia than FPG, as evidenced by recent studies showing that, in diabetic individuals, postmeal plasma glucose correlates better with GHb than does FPG (13)(14).
Biological variation has generally been defined as "random fluctuations around a homeostatic set-point" (9)(15). Several studies have examined biological variation in diabetic individuals and have concluded that there is a significant biological variation in GHb values that must be considered when test results are interpreted (5)(15). It is important to note that in diabetes patients, fluctuations in GHb concentrations are not random but are "pathologic", i.e., caused by changes in mean glycemia. It is also unclear how a homeostatic setpoint can be determined for an individual with diabetes because the setpoint itself can (and often does) change over time.
In summary, these data suggest that between-subject variation in GHb is minimal and is therefore not a major consideration when GHb is used for routine clinical care. Assay quality, however, is an important factor when interpreting GHb results (16)(17)(18), and imprecise assays may compromise the clinical utility of the test.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  C. D. Saudek, W. H. Herman, D. B. Sacks, R. M. Bergenstal, D. Edelman, and M. B. Davidson A New Look at Screening and Diagnosing Diabetes Mellitus J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2447 - 2453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W S A Smellie What is a significant difference between sequential laboratory results?Calf muscle pain can indicate localised vasculitis J. Clin. Pathol., April 1, 2008; 61(4): 419 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Cohen A1C: Does One Size Fit All? Diabetes Care, October 1, 2007; 30(10): 2756 - 2758. [Full Text] [PDF]
|
|
|
|
 |
|
 E. Selvin, C. M. Crainiceanu, F. L. Brancati, and J. Coresh Short-term Variability in Measures of Glycemia and Implications for the Classification of Diabetes Arch Intern Med, July 23, 2007; 167(14): 1545 - 1551. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Cohen, H. Snieder, C. J. Lindsell, H. Beyan, M. I. Hawa, S. Blinko, R. Edwards, T. D. Spector, and R. D. G. Leslie Evidence for Independent Heritability of the Glycation Gap (Glycosylation Gap) Fraction of HbA1c in Nondiabetic Twins Diabetes Care, August 1, 2006; 29(8): 1739 - 1743. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Droumaguet, B. Balkau, D. Simon, E. Caces, J. Tichet, M. A. Charles, E. Eschwege, and the DESIR Study Group Use of HbA1c in Predicting Progression to Diabetes in French Men and Women: Data from an Epidemiological Study on the Insulin Resistance Syndrome (DESIR) Diabetes Care, July 1, 2006; 29(7): 1619 - 1625. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. Saudek, R. L. Derr, and R. R. Kalyani Assessing Glycemia in Diabetes Using Self-monitoring Blood Glucose and Hemoglobin A1c JAMA, April 12, 2006; 295(14): 1688 - 1697. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. McCarter, J. M. Hempe, and S. A. Chalew Mean Blood Glucose and Biological Variation Have Greater Influence on HbA1c Levels Than Glucose Instability: An analysis of data from the Diabetes Control and Complications Trial Diabetes Care, February 1, 2006; 29(2): 352 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Williams, B. L. Cadwell, Y. J. Cheng, C. C. Cowie, E. W. Gregg, L. S. Geiss, M. M. Engelgau, K. M. V. Narayan, and G. Imperatore Prevalence of Impaired Fasting Glucose and Its Relationship With Cardiovascular Disease Risk Factors in US Adolescents, 1999-2000 Pediatrics, November 1, 2005; 116(5): 1122 - 1126. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Sacks and for the ADA/EASD/IDF Working Group of the HbA1c As Global Harmonization of Hemoglobin A1c Clin. Chem., April 1, 2005; 51(4): 681 - 683. [Full Text] [PDF]
|
|
|
|
 |
|
 E. Selvin, S. Marinopoulos, G. Berkenblit, T. Rami, F. L. Brancati, N. R. Powe, and S. H. Golden Meta-Analysis: Glycosylated Hemoglobin and Cardiovascular Disease in Diabetes Mellitus Ann Intern Med, September 21, 2004; 141(6): 421 - 431. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Goldstein, R. R. Little, R. A. Lorenz, J. I. Malone, D. Nathan, C. M. Peterson, and D. B. Sacks Tests of Glycemia in Diabetes Diabetes Care, July 1, 2004; 27(7): 1761 - 1773. [Full Text] [PDF]
|
|
|
|
|
|
R. J. McCarter, J. M. Hempe, R. Gomez, and S. A. Chalew Biological Variation in HbA1c Predicts Risk of Retinopathy and Nephropathy in Type 1 Diabetes Diabetes Care, June 1, 2004; 27(6): 1259 - 1264. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
