Diabetes Duration and Glycemic Control as Predictors of Cardiovascular Disease and Mortality

AIM The aim of this study was to assess the associations of diabetes duration and glycemic control (defined by plasma glycated hemoglobin [HbA1c] levels) with the risks of cardiovascular disease (CVD) and all-cause mortality and to determine whether the addition of either or both to the established CVD risk factors can improve predictions. METHODS A total of 435 679 participants from the UK Biobank without CVD at baseline were included. Cox models adjusting for classical risk factors (sociodemographic and anthropometric characteristics, lipid profiles and medication use) were used, and the predictive utility was determined by the C-index and net reclassification improvement (NRI). RESULTS Compared with participants without diabetes, participants with longer diabetes durations and poorer glycemic control had a higher risk of fatal/nonfatal CVD. Among participants with diabetes, the fully-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for diabetes durations of 5 to <10 y, 10 to <15 y and ≥ 15 y were 1.15 (0.99, 1.34), 1.50 (1.26, 1.79) and 2.22 (1.90, 2.58) (P-trend <0.01), respectively, compared with participants with diabetes durations <5 y; in addition, those with the longest disease duration (≥15 y) and poorer glycemic control (HbA1c ≥8% [53 mmol/mol]) had the highest risk of fatal/nonfatal CVD (HR 3.12; 95% CI 2.52, 3.86). Among participants with diabetes, the addition of both diabetes duration and glycemic control levels significantly improved both the C-index (change in C-index +0.0254; 95% CI 0.0111, 0.0398) and the overall NRI for fatal/nonfatal CVD (0.0992; 95% CI 0.0085, 0.1755) beyond the use of the classical risk factors. CONCLUSIONS Both a longer diabetes duration and poorer glycemic control were associated with elevated risks of CVD and mortality. Clinicians should consider not only glycemic control but also diabetes duration in CVD risk assessments for participants with diabetes. This article is protected by copyright. All rights reserved.