Haemoglobin A1c Time‐In‐Range and Mortality in Adults With Diabetes

ABSTRACT Introduction Glycemic variability over time is a risk factor for diabetes complications. We studied a new measure of A1c stability, A1c time in range and its association with mortality in adults with diabetes. Methods We conducted a retrospective cohort study of people with type 1 and type 2 diabetes from January 1, 2004 to December 31, 2018. Participants were 18 years or older with at least four A1c tests during a three‐year baseline. A1c time in range was calculated as the percentage of time during baseline when A1c levels were within person‐specific target ranges. We also calculated the percentage of time A1c levels were below target ranges to assess the effects of potential overtreatment. Adjusted models and instrumental variable models were used to measure associations between A1c time in range and mortality. Results We studied 84,229 participants with mean age 58.3 years (SD 12.1). In adjusted Cox models, each 20% decrement in A1c time in range was associated with increased mortality (A1c time in range [0% < 20%] HR, 1.30; 95% CI 1.23–1.37). Instrumental variable models to control for unmeasured confounding also showed associations between reduced A1c time in range (0% < 20%) and mortality (HR, 1.28, 95% CI 1.22–1.35). Mortality risks were similar in subgroups with or without prevalent cardiovascular disease and insulin users but were greater in older adults (≥ 60 years). Greater (≥ 60%) A1c time below range was associated with increased mortality (HR, 1.35, 95% CI 1.29–1.41). Conclusions Reduced A1c time in range and particularly greater time below range are associated with increased mortality in adults with diabetes. A1c stability within person‐specific ranges may be a risk factor for major outcomes in adults with diabetes.