US researchers have developed a more precise method for estimating average blood-sugar levels that can cut diagnostic errors by more than 50% compared with the current, widely used test. Practical Patient Care speaks to the senior investigator on the project, John Higgins of Harvard Medical School, about how he thinks this new method can lead towards a more convenient and correct form of patient diagnostics.
According to Dr John Higgins and his research, what we currently deem the gold standard for estimating average blood glucose is nowhere near as precise as it should be in order to accurately assess patients. This could lead to fatal errors; with an estimated global diabetic population currently spanning more than 422 million people, knowing accurate blood-sugar averages can help patients to better manage the disease and their risk of diabetes-related complications.
Together, the team of researchers at the prestigious Massachusetts General Hospital and Harvard Medical School developed a mathematical model to this end by integrating the mechanisms of haemoglobin glycation.
Practical Patient Care: Can you tell us a bit about your own background and your role in the study, especially the reasons you noticed this research needed undertaking?
Dr John Higgins: I am a pathologist at Massachusetts General Hospital in Boston, with expertise in computational and mathematical modelling. One of my interests is the population dynamics of circulating blood cells – how quickly we are producing cells of each type, how quickly they are maturing, how quickly we are clearing old ones, and how all of those rates vary from one person to the next or in one person over time under different conditions of health and disease.
We have accumulated evidence that red blood cell (RBC) lifespan is very tightly regulated in healthy individuals.Knowing how important HbA1c is in the management of diabetes, and how RBC lifespan will influence HbA1c independent of glucose levels, we decided to try to quantify the effect that RBC lifespan variation has on HbA1c.
I have been working with my colleague David Nathan, director of the Diabetes Center at Massachusetts General Hospital, for several years to try to answer this question.
Can you explain the research itself, the way it was undertaken and the results that were discovered?
We developed a mathematical model that integrated what we know about the chemistry of haemoglobin glycation and the biology of RBC kinetics. The model enabled us to take continuous glucose monitoring (CGM) measurements and a concurrent HbA1c, and calculate the magnitude of the non-glycaemic influence on HbA1c for each patient.
This patient-specific correction factor combined the effects of all hypothesised non-glycaemic factors (including RBC lifespan, RBC membrane glucose gradients, chemical glycation rates and reticulocytes glycation levels). We then noticed that the variation in this correction factor between patients was no more than the known variation in RBC lifespan between patients.
We could then conclude either that RBC lifespan is the only significant non-glycaemic factor causing glucose-independent variation in HbA1c, or that it is very highly correlated with other non-glycaemic factors – in which case, the other factors can be inferred from an estimate of a patient’s RBC lifespan.
Since we suspected from our prior research that RBC lifespan was very stable in each individual, we decided to try to estimate it and use it to correct each patient’s future HbA1c, attempting to provide a more accurate estimate of average glucose. We found that our method reduced errors in estimated average glucose by over 50% in four independent sets of more than 200 patients.
Why was this specific research undertaken? Why is a precise method for estimating average blood-sugar level a pressing matter, and how will it help the medical field?
HbA1c is the gold standard for diagnosing and managing diabetes, but we can make it much better and enable more precise disease management for individual patients if we take each patient’s RBC lifespan into account.
The research says that diagnostics errors are often caused by the AC1 test getting inaccurate samples from individual variations in the lifespan of a person’s RBCs. Can you explain this in more detail and why it’s so common?
RBC lifespan naturally varies from one person to the next. Other investigators have found that it can vary between 90 and 110 days in healthy people. Two people may have the same glucose level but, if their RBC lifespans differ by 20%, then the HbA1c will differ by 20%, and we will think their average glucose levels differ by that amount as well. The opposite can also happen, where RBC lifespan makes people with divergent glucose concentrations appear to have the same concentration.
Where do you think this leads diagnostics research in the hope for a more convenient and correct form of patient diagnostics testing?
We have shown that patients can get a significantly more accurate assessment of their disease management by adjusting their HbA1c level for their mean RBC age. Better assessment of current control will put patients in the position to reduce their risk of long-term disease complications. We’ve also shown that a patient’s mean RBC age can be estimated from a short period of CGM.
Will the more accurate test mean that patients have to have fewer of these tests, or that hospitals might be able to perform fewer of them?
I think patients will still benefit from frequent tests but, with a clearer report on how they are doing, they will be able to adjust their treatment more precisely.
What are your thoughts on the current state of pathology and diagnostics – especially with regard to where improvements can be made in patient care?
While we do show that the current use of the HbA1c can be significantly improved, it is also important to note that it represents a major advance on random glucose testing.
Otherwise, there is a lot of discussion about how big data is going to lead to advances in health care, and I think our study shows one way for actual improvements to be achieved: by thinking about physiologic mechanisms in a quantitative way, we can use the huge amount of data collected by continuous glucose monitors to reveal new details about physiology, and use that new understanding to manage each patient’s disease precisely.