Home / Therapies / Blood Glucose Control / Comparing Long-Term HbA1C Variability with All-Cause Mortality in the ACCORD Trial

Comparing Long-Term HbA1C Variability with All-Cause Mortality in the ACCORD Trial

Jun 16, 2020
Editor: David L. Joffe, BSPharm, CDE, FACA

Author: Stephen Rubano, PharmD. Candidate, USF Taneja College of Pharmacy

Study shows that visit-to-visit long-term HbA1C variability can have significant effects on the prognosis for patients living with diabetes. 

Glycemic control is the priority of any treatment plan addressing diabetes, with goals set by the American Diabetes Association (ADA) of an HbA1c level of below seven percent. Substantial evidence exists supporting the claims that elevated HbA1c levels increase the risk of micro- and macrovascular complications in patients with diabetes. Because of this, intensive treatment plans aimed at restoring normal HbA1c levels as quickly as possible have been tested. However, several theories suggest that while this goal is essential, the process used in reaching the goal is equally important. For example, the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial determined that high-intensity treatments showed higher all-cause mortality rates than standard treatment options. These findings raised the alarm among researchers leading to the premature termination of the study. More focus has been placed on studying the effects of long-term HbA1c variability as a predictor of mortality over average HbA1c reduction. Previous studies show that higher degrees of HbA1c variability are associated with increased all-cause mortality and worse patient outcomes. However, many of these studies are limited by insufficient parameters such as small sample size or inadequate glycemic measuring. 


To better understand this phenomenon, researchers performed a post-hoc analysis using results from the ACCORD trial to find whether HbA1c variability offers any prognostic significance. In efforts to quantify results, the study defined variability as the difference between visit-to-visit HbA1c levels taken every four months among participants. Furthermore, analysis of variability was expressed using three separate indices, including coefficient of variation (CV), variability independent of the mean (VIM), and average real variability (ARV). The study used statistical software to test for significance with a standard two-tail alpha level of 0.05. Long-term HbA1c variability assessment included three or more HbA1c measurements retrieved from a total of 10,251 patients. Association between the variables of interest and mortality were estimated using multivariable Cox proportional hazards models adjusting for potential confounders such as sex, medical history, age, BMI, among other factors.  

Major baseline characteristics showed that participants had an average age of 62.7 years and were primarily male (61.8%). Researchers uncovered several key findings, most notably, all three variable indices (CV, VIM, ARV) were significantly associated with all-cause mortality (P < 0.01) regardless of the treatment group. For both the standard and intensive treatment groups, hazard ratios (HRs) were calculated for increasing HbA1c variability and found to be 1.50 and 1.28, respectively. The cross-tabulation analysis revealed that both increasing HbA1c means and HbA1c variability were linked with more significant all-cause mortality. Although increases in both variables resulted in a higher risk of mortality, mean and variability remained independent in terms of the effect on outcomes. Finally, the ARV of HbA1c exhibited higher HRs in the intensive treatment group than the standard treatment group likely due to more pronounced HbA1c fluctuations. 

The post-hoc analysis of the ACCORD trial provides evidence on two key issues in treating patients with diabetes. First, long-term visit-to-visit HbA1c variability is a powerful predictor of all-cause mortality, even after accounting for mean HbA1c levels. Second, increases in both HbA1c variability and average HbA1c levels are associated with a higher risk of mortality. Results should be interpreted considering the study’s limitations, one of which is HbA1c was the primary outcome target rather than glucose levels. However, these findings are consistent with previous observational research. For example, two previous prospective cohort studies, conducted in 2012 and 2015, showed an increased risk of diabetic neuropathy in patients with higher HbA1c variation. While the reduction in HbA1c levels remains a focus in diabetic treatment, achieving glucose stability within the patient plays a crucial role in success. Identifying the root causes of HbA1c variation and reducing variability can improve overall mortality for patients living with diabetes.  

Practice Pearls:  

  • Long-term HbA1c variability has been linked to an increased risk of all-cause mortality in patients with diabetes. 
  • A post-hoc analysis of the ACCORD trial shows the risks of intensive treatment and the importance of maintaining stable glucose levels. 
  • Understanding the causes of HbA1c variability within patients can lead to better health outcomes and reduce the risk of complications. 


Sheng, Chang-Sheng, et al. Prognostic Significance of Long-Term HbA1c Variability for All-Cause Mortality in the ACCORD Trial. Diabetes Care, vol. 43, no. 6, 2020, pp. 1185–90.  doi:10.2337/dc19-2589. 

Rodríguez-Segade, et al. Intrapersonal HbA1cvariability and the Risk of Progression of Nephropathy in Patients with Type 2 Diabetes. Diabetic Medicine, vol. 29, no. 12, 2012, pp. 1562–66.  doi:10.1111/j.1464-5491.2012.03767.x. 


Stephen Rubano, PharmD. Candidate, USF Taneja College of Pharmacy 


See more about HbA1c variability.