Differences Between Glycemia Estimates from Hemoglobin A1c and Continuous Glucose Monitoring and Their Association with Complete Blood Counts

Authors

Veronica Tozzo, Department of Pathology and Center for Systems Biology, Massachusetts General Hospital, Boston, Massachusetts, USA.
David M. Nathan, Diabetes Research Center, Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts, USA.
Heidi Krause-Steinrauf, The Biostatistics Center, Department of Biostatistics and Bioinformatics, Milken Institute School of Public Health, The George Washington University, Rockville, Maryland, USA.
John M. Lachin, The Biostatistics Center, Department of Biostatistics and Bioinformatics, Milken Institute School of Public Health, The George Washington University, Rockville, Maryland, USA.
Christopher Mow, Department of Pathology and Center for Systems Biology, Massachusetts General Hospital, Boston, Massachusetts, USA.
Nicole Butera, The Biostatistics Center, Department of Biostatistics and Bioinformatics, Milken Institute School of Public Health, The George Washington University, Rockville, Maryland, USA.
Robert M. Cohen, Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, University of Cincinnati, Cincinnati, Ohio, USA.
John M. Higgins, Department of Pathology and Center for Systems Biology, Massachusetts General Hospital, Boston, Massachusetts, USA.

Document Type

Journal Article

Publication Date

12-15-2025

Journal

Diabetes technology & therapeutics

DOI

10.1177/15209156251395036

Keywords

complete blood count; continuous glucose monitoring; hemoglobin A1c; hemoglobin A1c-glucose discordance; type 2 diabetes

Abstract

Continuous glucose monitoring (CGM) and hemoglobin A1c (HbA1c) provide estimates of mean glycemia that may differ, in part, due to the effects of variation in red blood cell (RBC) age and turnover on HbA1c. Measurements derived from the complete blood count (CBC) may vary with RBC age and might be used to reduce the difference between glycemia estimates derived from CGM and HbA1c. We analyzed CBC measurements from 1,325 individuals with type 2 diabetes who participated in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) CGM substudy. Mean glycemia was estimated from HbA1c (eAG) using the A1c-Derived Average Glucose (ADAG) formula and from CGM by averaging 10 days of measurements (eAG). We evaluated the association between CBC-derived data and the difference (eAG - eAG) using linear models, both unadjusted and adjusted for age and self-identified sex. In adjusted analyses, several CBC-derived measurements were significantly associated with the difference between eAG and eAG. Platelet count and RBC distribution width (RDW) were positively associated, while hemoglobin concentration (HGB), reticulocyte fraction, mean corpuscular volume (MCV), mean corpuscular hemoglobin content (MCH), mean corpuscular hemoglobin concentration (MCHC), and reticulocyte MCHC were negatively associated. A linear model from HbA1c to eAG adjusted with all significantly associated CBC measurements (CBC-AG) provided modestly improved estimates of eAG compared with ADAG, with R (SD) for ADAG of 0.68 (0.07) and for CBC-AG 0.72 (0.06). CBC measurements are associated with differences between estimates of glycemia derived from HbA1c and CGM. Further studies with longer periods of CGM are needed to determine whether CBCs can complement HbA1c and CGM and can help reconcile differences in estimates of mean glycemia provided by HbA1c and CGM.

APA Citation

Tozzo, Veronica; Nathan, David M.; Krause-Steinrauf, Heidi; Lachin, John M.; Mow, Christopher; Butera, Nicole; Cohen, Robert M.; and Higgins, John M., "Differences Between Glycemia Estimates from Hemoglobin A1c and Continuous Glucose Monitoring and Their Association with Complete Blood Counts" (2025). GW Authored Works. Paper 8398.
https://hsrc.himmelfarb.gwu.edu/gwhpubs/8398

Department

Biostatistics and Bioinformatics