Estimation of hemoglobin A1C using the complete blood count measures in the diagnosis of diabetes

Abstract

Objective: Diabetes is a metabolic disorder occurring either due to the inadequate secretion of insulin or ineffective utilization of insulin by the body. The study was aimed to identify the variations of the complete blood count (CBC) parameters among the diabetic and normal individuals and to derive an empirical formula to estimate hemoglobin A1c (HbA1c) of an individual using CBC parameters. Methods: A total of 83 subjects (mean age: 52.8±9.0 years) involved in the study, among which 39 (mean age: 49.1±8.8 years) were normal and 44 (mean age: 56±7.8 years) were diabetic. The blood was drawn from the participants and was subjected to CBC analysis using automated hematology analyzer. The stepwise linear regression model was used to determine the empirical formula to estimate HbA1c using the CBC parameters. The Student’s t-test was performed to identify the group differences. Results: A negative correlation was observed for Hb (r=−0.35**, p<0.001) and packed cell volume (PCV) (r=−0.23**, p<0.05) against HbA1c. The CBC parameters Hb, erythrocyte sedimentation rate, PCV, red blood cells count, mean corpuscular volume, and mean corpuscular Hb exhibited a statistically significant difference at the level (p<0.05) between the normal and diabetic groups. The empirically derived formula yielded sensitivity, specificity, positive predictive value, negative predictive value, and accuracy measures of 91%, 49%, 67%, 83%, and 71%, respectively, in diagnosing diabetes based on the estimated HbA1c. Conclusion: The empirical formula derived to estimate HbA1c could be useful in the prediction of diabetes with an appreciable accuracy.