Department of Biostatistics Seminar/Workshop Series

From discordance to discovery in diabetes and beyond

Christopher J. Lindsell PhD, Professor and Vice Chair of Research, Department of Emergency Medicine, University of Cincinnati Medical Center, Vice President for Research, UC Health, Associate Dean for Clinical Research, College of Medicine, University of Cincinnati

Some persons with diabetes have good glucose control but high HbA1c while others have poor glucose control but a normal HbA1c. While the implications of this observed discordance were initially unclear, our studies confirmed that it was not random. The discordance was repeatable and, using a quantitative genetic model in a classic twin study, we found that it was a heritable trait. The discordance was also associated with the development of retinopathy, hinting at both risk prediction and pathophysiology. To understand the nature of the discordance, we considered variables that might result in differences between extra-cellular measures of glucose control and the HbA1c measured in the red blood cell. One possibility was red blood cell lifespan; the longer the red blood cell survives, the greater the glycation of hemoglobin. Tracking survival of red blood cells labeled with biotin, we observed that inter-individual variability in red blood cell survival is of sufficient magnitude to alter interpretation of the HbA1c. Using a mathematical association between the survival curve and the age distribution of the random sample of red blood cells, we demonstrated that correcting HbA1c for red blood cell age improves the accuracy of the association between HbA1c and mean plasma glucose. While this finding has potential clinical utility for the diagnosis and management of diabetes, the biotin method is impractical for the clinical setting. We therefore validated a safe, accurate and low cost method for measuring red blood cell lifespan using an oral stable isotope, ¹⁵N-glycine. As well as successfully replicating our prior studies, we have now used this methodology to accurately measure red blood cell survival in sickle cell and demonstrated that commonly used surrogate measures of hemolysis are inaccurate. This has implications for studies that have relied on these surrogate markers. In summary, drawing inferences from discordance between two variables commonly considered equivalent has provided biological insights in diabetes and beyond.