Relationship of HbA1c variability, absolute changes in HbA1c, and all-cause mortality in type 2 diabetes: a Danish population-based prospective observational study

OBJECTIVE We assessed the relationship of mortality with glycated hemoglobin (HbA1c) variability and with absolute change in HbA1c. DESIGN A population-based prospective observational study with a median follow-up time of 6 years. METHODS Based on a validated algorithm, 11 205 Danish individuals with type 2 diabetes during 2001-2006 were identified from public data files, with at least three HbA1c measurements: one index measure, one closing measure 22-26 months later, and one measurement in-between. Medium index HbA1c was 7.3%, median age was 63.9 years, and 48% were women. HbA1c variability was defined as the mean absolute residual around the line connecting index value with closing value. Cox proportional hazard models with restricted cubic splines were used, with all-cause mortality as the outcome. RESULTS Variability between 0 and 0.5 HbA1c percentage point was not associated with mortality, but for index HbA1c ≤8% (64 mmol/mol), a variability above 0.5 was associated with increased mortality (HR of 1 HbA1c percentage point variability was 1.3 (95% CI 1.1 to 1.5) for index HbA1c 6.6-7.4%). For index HbA1c≤8%, mortality increased when HbA1c declined, but was stable when HbA1c rose. For index HbA1c>8%, change in HbA1c was associated with mortality, with the lowest mortality for greatest decline (HR=0.9 (95% CI 0.80 to 0.98) for a 2-percentage point decrease). CONCLUSIONS For individuals with an index HbA1c below 8%, both high HbA1c variability and a decline in HbA1c were associated with increased mortality. For individuals with index HbA1c above 8%, change in HbA1c was associated with mortality, whereas variability was not.