Eight-year HbA1c burden and incident cardiovascular disease after accounting for the most recent HbA1c in older US adults: a cohort study

Background: The most recent glycated haemoglobin (HbA1c) measurement reflects current glycaemic status but does not capture preceding glycaemic exposure. We examined whether, among older US adults with the same most recent HbA1c level, a greater eight-year, time-weighted HbA1c burden derived from three serial measurements remained associated with incident cardiovascular disease (CVD). Methods: We conducted an observational cohort study using the Health and Retirement Study. Eligible participants had three HbA1c measurements spanning eight years and were free of self-reported heart disease and definite stroke at a 2014 or 2016 landmark. HbA1c burden was calculated as the trapezoidal time-weighted mean of the three measurements. Incident CVD was the first post-landmark report of heart disease or definite stroke over three biennial intervals. Pooled complementary log-log models estimated cause-specific hazard ratios (HRs) and 95% confidence intervals (CIs) for HbA1c burden, with simultaneous adjustment for the most recent HbA1c, hypertension, diabetes, and prespecified covariates. Death before CVD was treated as a competing event. Sensitivity analyses included a two-year lag, multiple imputation, and alternative handling of same-interval CVD and death. Results: Among 3,377 participants (mean age, 71.9 years; 64.2% women), 589 incident CVD events occurred and 332 participants died before CVD. Each 1-percentage-point higher HbA1c burden was associated with a 33% higher cause-specific CVD hazard (HR, 1.33; 95% CI, 1.11–1.59; P = .002). For the most recent HbA1c, the conditional HR was 0.90 (95% CI, 0.78–1.05; P = .189). Estimates were similar in the two-year lag analysis (HR, 1.32; 95% CI, 1.05–1.66) and after multiple imputation (HR, 1.35; 95% CI, 1.13–1.62). When 87 same-interval CVD/death records were classified as competing deaths or excluded, the HRs were 1.19 (95% CI, 0.98–1.46) and 1.21 (95% CI, 0.99–1.48), respectively. Conclusions: In this cohort of older US adults, higher eight-year HbA1c burden was associated with incident CVD after accounting for the most recent HbA1c in the primary analysis. These findings support further investigation of glycaemic history in cardiovascular risk assessment.