Dysglycaemia and the risk of acute myocardial infarction in multiple ethnic groups: an analysis of 15,780 patients from the INTERHEART study

Abstract

Aims/hypothesis: Although diabetes is an established risk factor for myocardial infarction (MI), disease control may vary. HbA(1c) is a reliable index of ambient glucose levels and may provide more information on MI risk than diabetes status.

Methods: The relationship between HbA(1c) levels in MI patients and controls who participated in the 52 country INTERHEART study was analysed.

Results: In 15,780 participants with a HbA(1c) value (1,993 of whom had diabetes), the mean (SD) levels for HbA(1c) were 6.15% (1.10) in the 6,761 MI patients and 5.85% (0.80) in the control participants. After adjustment for age, sex and nine major MI risk factors (including diabetes), higher HbA(1c) fifths above the lowest fifth (HbA(1c) <5.4%) were associated with progressively higher OR of MI, with OR for the highest HbA(1c) fifth (≥ 6.12%) being 1.55 (95% CI 1.37-1.75). When analysed as a continuous variable after adjustment for the same factors, every 1% higher HbA(1c) value was associated with 19% (95% CI 14-23) higher odds of MI, while every 0.5% higher HbA(1c) was associated with 9% higher odds of MI (95% CI 7-11). Concordant relationships were noted across subgroups, with a higher OR noted in younger people, patients without diabetes or hypertension, and those from some regions and ethnicities.

Conclusions/interpretation: The HbA(1c) value provides more information on MI odds than self-reported diabetes status or many other established risk factors. Every 1% increment independently predicts a 19% higher odds of MI after accounting for other MI risk factors including diabetes.