Glycosylated hemoglobin, or specifically, hemoglobin A1c (HbA1c), is a marker of average blood-glucose concentrations over a two- to three-month period. In the first study, researchers, with first author Dr Kay-Tee Khaw (University of Cambridge School of Clinical Medicine, UK), looked at the relationship between HbA1c, cardiovascular disease, and total mortality in subjects enrolled in the European Prospective Investigation into Cancer in Norfolk (EPIC-Norfolk) study, a prospective population study.

Blood-glucose thresholds diagnostic of diabetes have been based on the occurrence of microvascular complications, they write. But for macrovascular complications such as CV disease, increasing evidence suggests the relationship between blood glucose and risk is continuous, with no obvious threshold, Khaw and colleagues write.

In a previous report from this same cohort,4 they had shown that HbA1c levels were related to CV risk and all-cause mortality but didn't have sufficient power to look at the relationship with CV disease at concentrations close to the diagnostic threshold of 7% or to examine the relationship in women, the authors note.

In this paper, HbA1c and CV risk factors were assessed in 4662 men and 5570 women between 1995 and 1997, and they were followed out to 2003. Average follow-up was six years.

They found that the relationship between HbA1c and both CV disease and all-cause mortality was "continuous and significant throughout the whole distribution" and was apparent even among those without known diabetes. An increase in HbA1c of 1% was associated with a 24% increase in the relative risk for all-cause mortality in men and a 28% increase in women, independent of other CV risk factors.

When those with known diabetes, HbA1c levels of 7% or greater, or known CV disease were excluded, the result was largely the same (RR 1.26 [95% CI 1.04-1.52]; p=0.02).

Most of the events occurred in those with moderately increased HbA1c levels: 72% of deaths occurred in those with HbA1c concentrations between 5% and 6.9%, compared with 15% among those with frank diabetes. In their discussion, the researchers point out that the increased risk associated with diabetes may be mediated "almost entirely" through HbA1c concentrationdiabetes was no longer significant when HbA1c was added to the regression model, they point out.

Although they acknowledge their findings don't establish a causal relationship, Khaw et al conclude, "Our study highlights the independent relationship of blood-glucose concentrations to cardiovascular and mortality risk in men and women in the general population and supports the need for randomized, controlled trials of interventions to assess the effect of reduced hemoglobin A1c concentrations on cardiovascular events in persons without diabetes."

In a separate report in the same issue, Elizabeth Selvin (Johns Hopkins University, Welch Center for Prevention, Epidemiology and Clinical Research, Baltimore, MD) and colleagues report results of a meta-analysis, examining the relationship between glycosylated hemoglobin and CVD in people with diabetes.

Selvin et al also pointed out that chronic hyperglycemia, assessed by glycosylated hemoglobin level, has clearly been related to the risk of microvascular complications, but the relationship with macrovascular complications is less clear.

This report is a meta-analysis of prospective cohort studies in people with diabetes that included data on glycosylated hemoglobin levels and on incident CVD. They included three studies of 1688 subjects with type 1 diabetes and 10 studies of 7435 subjects with type 2 diabetes.

The pooled relative risk for total CVD in those with type 2 diabetes was 1.18 for each 1-percentage point increase in glycosylated hemoglobin (95% CI 1.10-1.26). For coronary heart disease, an end point examined in five of these studies, the pooled relative risk was 1.13 (95% CI 1.06-1.20) for each 1-percentage point increase in glycosylated hemoglobin, and for fatal CHD, it was 1.16 (95% CI 1.07-1.26).

Results in the studies of type 1 diabetes were similar, they write, but with wider confidence intervals. The pooled relative risk was 1.15 (95% CI 0.92-1.43) for each 1-percentage point increase in glycosylated hemoglobin.

Pending confirmation from large ongoing clinical trials, Selvin et al conclude that these findings suggest improvements in glycemic control may lower the risk for CVD in persons with diabetes. Because CVD is the most common cause of death in this population, they add, "additional studies are needed to decipher the independent effect of glycosylated hemoglobin in predicting cardiovascular disease outcomes in persons with diabetes."