LAUSR

Introduction: HbA1c is central to diagnosis and management of diabetes. However, our understanding of the associations of HbA1c with clinical outcomes is based primarily on studies of middle-aged adults. Objective: To characterize the associations between HbA1c and mortality among older adults with and without diabetes. We compared the HbA1c-mortality associations with those for alternative glycemic markers (fructosamine and glycated albumin). Methods: We conducted a prospective cohort analysis of 6370 participants (32% with diagnosed diabetes, mean age 76, 59% female, 23% black) in the Atherosclerosis Risk in Communities (ARIC) Study, baseline visit 5 (2011-13). We used Cox regression models to examine the association of each glycemic biomarker (modeled as a linear spline) with all-cause mortality through 2017, stratified by diagnosed diabetes. Model discrimination was tested using c-statistics. Results: There were 1022 deaths over 6 years of follow-up. In persons with diabetes, there was a J-shaped association between HbA1c and mortality ( Figure ). Associations were largely similar for fructosamine and glycated albumin. Among persons without diabetes, HbA1c was not strongly associated with mortality, but confidence intervals were wide for HbA1c <5.0% and ≥6.5% (undiagnosed diabetes); whereas, higher levels of fructosamine and glycated albumin were associated with an elevated mortality risk. Associations were similar before and after adjustment for cardiovascular risk factors. The addition of individual glycemic markers to the models modestly improved discrimination in both persons with and without diabetes. C-statistics were similar when the glycemic markers were added individually to the diabetes-stratified models (Figure). Conclusions: Among older adults, HbA1c and other glycemic markers tend to have similar associations with mortality. HbA1c, fructosamine, and glycated albumin, including low levels, may reflect common markers of risk in older adults, particularly in those with diabetes.