Association between HbA _1c and cardiovascular disease mortality in older Hong Kong Chinese with diabetes

Objective To examine the association between baseline HbA _1c level and mortality attributable to all-cause, cardiovascular disease, coronary heart disease and stroke. Methods A total of 2137 Chinese aged 65years or above attending the Elderly Health Service, Department of Health, Hong Kong, with diagnosed diabetes had HbA _1c measured during 1998 to 2000 and were followed up to 2009. Information on socio-economic position, lifestyle factors and disease history was collected. Hazard ratios and 95% confidence intervals with adjustment for potential confounders were calculated using Cox's proportional hazards models. Results After an average of 7.9years of follow-up, 540 participants had died. After adjusting for potential confounders, higher HbA _1c (≥69mmol/mol, 8.5%) increased the risk of cardiovascular disease (hazard ratio2.11;95%CI 1.37-3.25) and stroke mortality (hazard ratio2.43; 95%CI 1.06-5.55) compared with HbA _1c of 58-68mmol/mol (7.5-8.4%), and increased the risk of all-cause (hazard ratio1.41; 95%CI 1.06-1.86) and coronary heart disease mortality (hazard ratio2.44; 95%CI 1.11-5.37) compared with HbA _1c of 48mmol/mol (6.5%) or less. Analysis of HbA _1c as a continuous variable showed that every XXmmol/mol (1%) increase in HbA _1c decreased stroke mortality risk by 51% in those with HbA _1c level less than 48mmol/mol (6.5%) and increased stroke mortality risk by 30% in those with an HbA _1c level of 48mmol/mol (6.5%) or higher, suggesting a U-shaped association between HbA _1c and stroke mortality. Conclusion High HbA _1c predicted excess risk of all-cause, cardiovascular disease, coronary heart disease and stroke mortality. The question of whether low HbA _1c increases mortality in older patients with diabetes needs further investigation.