Expert Blog

When Does Tight Glycemic Control Affect CVD Risk?

Carol Wysham, MD
Endocrinologist

It is abundantly clear that cardiovascular disease is a major driver of the morbidity, mortality, and health care costs associated with type 2 diabetes mellitus (T2DM). 

Whereas multiple short-term clinical trials have demonstrated reduction of CV events with control of blood pressure and LDL-cholesterol, three short-term trials (ACCORD, ADVANCE, and VADT) failed to show a reduction in cardiovascular events with tight glycemic control; in fact there was higher rate of death in the intensively treated patients in ACCORD.  There is little doubt that abnormal glucose regulation plays a major role in the excess cardiovascular risk in diabetes. Why, then, did these trials fail to show benefit of tight glycemic control?  Is it possible that the effects of tight glycemic control vary according to when, in the course of the disease, it is instituted?

If initiated after several years of uncontrolled dysglycemia, the risk of tight glycemic control may exceed the benefit. Autopsy series have shown that the earliest phase of vascular disease begins many decades before the appearance of clinical events. Studies show that hyperglycemia is responsible for the initial appearance of fatty streaks in the vessels of diabetic animals, through mechanisms such as induction of increased levels of oxidative stress, inflammatory cytokines, adhesion molecules, and glycosylation of proteins.  Hyperglycemia continues to play a role as these fatty streaks progress over time to atherosclerotic plaque. Significantly, correction of hyperglycemia with insulin therapy can prevent these early vascular changes. However, correction of hyperglycemia later in the course of disease is unlikely to cause regression of established plaque. Add to that the fact that, as beta cell function and mass decline over time, tight glycemic control becomes much more difficult. As glucose levels become more variable, the risk for hypoglycemia increases, possibly triggering acute cardiovascular events through hemodynamic, rheologic, and electrolyte alterations in those with established vascular disease.

Although the official analyses of the ACCORD trial did not find a relationship between hypoglycemia and increased mortality, the protocol was not designed to capture hypoglycemia accurately. The case report forms only recorded information about severe or symptomatic hypoglycemia. Episodes of biochemical hypoglycemia were captured only if  they occurred within the two weeks prior to the research visit. Many of the intensively treated patients at our center experienced frequent episodes of hypoglycemia, with SMBG values below 55 mg/dl. Commonly, these episodes were not associated with symptoms (due to hypoglycemia unawareness) and occurred outside of the recording interval. Therefore, the data from the ACCORD trial could  not be used to rule out the association between mild to moderate hypoglycemia and risk for CV events.

That early, aggressive antihyperglycemic therapy might be associated with improved cardiovascular outcomes can be inferred from subanalyses of ACCORD and VADT, as well as from longer-term glycemia trials (UKPDS and DCCT). In the ACCORD trial, intensive glycemic control appeared to be associated with fewer CV events in patients with no history of cardiovascular disease and in those with baseline A1c < 8%. Similarly, in the VADT, there appeared to be lower risk for CV events in intensively treated subjects with a < 15 year history of diabetes and in those with coronary calcium score < 100. 

Finally, the UKPDS and DCCT failed to demonstrate a benefit of tight glycemic control on CV events in newly diagnosed subjects at the end of the intervention, in large part due to low numbers of events.  However, both studies demonstrated a reduction in CV events in follow-up ten years after the end of the intervention, despite similar glycemic control in the post-intervention period.  These findings have been attributed to deleterious effects of hyperglycemia that are not reversible and further support the importance of early intervention.

Prevention of the cardiovascular complications of diabetes requires aggressive control of all modifiable risk factors. Optimally, people with diabetes would benefit from early diagnosis and aggressive glycemic control from the time of diagnosis. In the presence of diabetes of long duration, especially with known vascular disease, however, the risks of very intensive glycemic control might outweigh the risks. In these individuals, higher targets might be necessary to ensure patient safety.