Glycated haemoglobin and the association with long-term survival among patients evaluated for stable angina – a prospective Norwegian cohort

Abstract Background and methods Recent guidelines have included glycated haemoglobin (HbA1c) ≥48 mmol/L as a diagnostic criterion for diabetes mellitus (DM) in addition to plasma glucose (PG) concentrations, mainly based on the relationship between hyperglycemia and microvascular disease [1]. However, increased HbA1c may stem not only from hyperglycemia, and the risk association between HbA1c and long-term survival in patients with stable coronary heart disease and HbA1c ≥48 mmol/L but no previous DM according to PG is uncertain. We explored the relationship between HbA1c and survival among patients with and without DM who were evaluated for stable angina in the period 2000–2004. Endpoints were obtained from the Norwegian Cause of Death Registry. Results In total, 4164 patients were evaluated by cardiac cathetherization, of whom 576 patients (13.8%) had DM (median HbA1c 55 mmol/L) according to self-report and/or baseline PG concentrations. Of the remaining 3588 patients 1026 had HbA1c ≥48 mmol/L; however, HbA1c did not correlate with the HOMA2 insulin resistance index or fasting PG in these patients. During median (25–75 percentile) follow-up time of 14.0 (12.1–15.4) years a total of 1328 patients (31.9%) died, of whom 582 from cardiovascular causes. In patients with DM according to PG, HbA1c trended towards positive associations with all-cause and CVD mortality when adjusted for age and gender (HRs (95% CIs) 1.13 (0.99–1.28) and 1.16 (0.98–1.39) per 1SD, respectively). However, HbA1c was not associated with survival in either the group of patients without DM and HbA1c <48 mmol/L (median HbA1c 38 mmol/L) (HRs (95% CIs) 0.99 (0.92–1.06) and 0.96 (0.86–1.08) for all-cause and CVD mortality, respectively) or patients without DM but having HbA1c ≥48 mmol/L (median HbA1c 53 mmol/L) (HRs (95% CIs) 0.99 (0.88–1.12) and 1.04 (0.88–1.22)). Conclusion In patients evaluated for stable angina pectoris about two decades ago, almost a third of patients with no history of DM according to PG still had HbA1c concentrations indicating DM according to current guidelines. Including these patients in the DM category yielded similar percentages of patients with DM as observed in recent populations with stable coronary disease [2]. However, as opposed to what we observed in patients with DM, HbA1c did not show any association with very long-term survival among patients without DM. Our findings therefore question the use of HbA1c in the diagnosis of DM, especially in terms of risk assessment for longevity among patients with chronic coronary syndrome. Funding Acknowledgement Type of funding sources: None.