Glycated hemoglobin variability: A potential new risk marker for diabetes complications?

Abstract

Increasing evidence has shown that glycemic instability might contribute to the development of diabetes complications, in addition to the strong relationship between elevated glycemia and diabetes complications risk. In a recent issue of Diabetes Care, an analysis of The Renal Insufficiency And Cardiovascular Events (RIACE) Italian Multicenter Study by Penno et al.1 has added to the evidence that variability in glycated hemoglobin (HbA1c) affects albuminuria and albuminuric chronic kidney disease (CKD) phenotypes independently of, or instead of, average HbA1c. On the contrary, diabetic retinopathy mainly depends on the average HbA1c, not variability1. Glucose variability refers to multiple fluctuations of glycemia in the same individual within-day or day-to-day, or even over longer periods of time; that is, week-to-week or visit-to-visit. The concept of glucose variability was first introduced in the Diabetes Control and Complications Trial (DCCT), and defined as the standard deviation (SD) of daily blood glucose around the mean from each quarterly visit, even though it proved a weak relationship between glucose variability and diabetes complications2. In comparison with the inconsistent association between pre- or postprandial glucose variability and microvascular complication risk, HbA1c variability from visit-to-visit was found to be an independent risk factor for the development of diabetes complications in type 1 diabetes, either solely or in addition to the effect of average HbA1c. In the current RIACE study, Penno et al.1 retrospectively addressed the associations between average HbA1c and HbA1c variability with microvascular complications in Caucasian participants with type 2 diabetes, covering the entire spectrum of renal disease and diabetic retinopathy (Table 1). The RIACE cohort consisted of 15,933 patients from 19 hospital-based diabetes clinics throughout Italy, with three to five HbA1c values in serial visits within a 2-year period before enrolment. In accordance with Tsukuba Kawai Diabetes Registry 2 in Japan3 and the Diabetes Management through an Integrated Delivery System project in Taiwan4, the RIACE study added to the fact that HbA1c variability was independently correlated with microalbuminuria and early stages of CKD to the same extent as average HbA1c in patients with type 2 diabetes. Furthermore, they provided the first evidence that HbA1c variability exclusively correlated with macroalbuminuria and the progressive stages of albuminuric CKD instead of average HbA1c. Even though there was substantial data regarding the effect of HbA1c variability on nephropathy risk and the magnitude varied in these studies, the RIACE highlighted the predominant role of HbA1c variability over average HbA1c in accelerating the development of albuminuric nephropathy and reduced estimated glomerular filtration rate. Patients with more variable HbA1c face a higher risk of microvascular complications, in terms of the frequency and amplitude of HbA1c fluctuation. Previous scientific research attributed the deleterious effect of glucose variability on the kidneys to the metabolic memory induced by repeated exposure to glucose fluctuation. The precise mechanism has not been well determined; however, endothelial dysfunction and oxidative stress were found to be worsened by glucose variability compared with stable hyperglycemia, and could be reversed by reduction of glucose fluctuation5. The RIACE cohort especially emphasized the effect of amplitude of HbA1c fluctuation, as shown by the fact that the risk of progressive nephropathy increased significantly with a higher quartile of HbA1c SD rather than a lower quartile of HbA1c SD1. Thus, patients lagged in the ‘metabolic memory’ as a result of frequent HbA1c fluctuation with a large range, and were much more prone to developing severe nephropathy than those with the same average HbA1c, but less variable HbA1c. In contrast to nephropathy, the association between retinopathy and HbA1c variability was less consistent. Penno et al.1 speculated that the magnitude of HbA1c variability was not high enough to affect the development of diabetic retinopathy, and was masked by average HbA1c and other possible variables related to glycemic exposure, such as diabetes duration and treatments1. One fact we cannot deny is that type 1 diabetes might have more variable glycemia than type 2 diabetes does. Furthermore, the RIACE cohort with type 2 diabetes showed that increasing average HbA1c was associated with longer diabetes duration, whereas higher HbA1c SD was linked to shorter diabetes duration. In fact, diabetes duration is considered to be the probable strongest predictor for development and progression of retinopathy. With respect to these facts, there raises the possibility of whether risk reduction of diabetes complications could be achieved by reducing variable HbA1c. Ideal interventions to hyperglycemia aim at lowering hyperglycemia without running the risk of hypoglycemia. Now it might provide benefit beyond simply reducing the risk of hypoglycemia, by avoiding the risk of diabetes complications. To distinguish whether any positive outcomes are as a result of the reduced glycemic variability or just the proved glycemic reduction is still challenging. At least not all diabetes complications are associated with HbA1c variability. More powerful interventional trials should aim to address these remaining questions.