Diabetes study helps clear up contradictions on treatment

Times Staff Writer

Lowering blood sugar levels to near normal through intensive treatment may reduce the risk of cardiovascular disease and heart attacks for Type 2 diabetics, but only if treatment is begun relatively soon after diagnosis and if severe episodes of low blood sugar can be avoided, VA researchers reported Sunday.

The findings from the Veterans Affairs Diabetes Trial may help to resolve concerns raised by two widely reported studies: One found no apparent benefit from such treatment, and the other concluded that intensive treatment might actually be detrimental.

Data from all three studies were presented over the weekend at a San Francisco meeting of the American Diabetes Assn., and although many questions remain unanswered, some generalizations are becoming possible.

Foremost, data from all three studies suggest that the greatest reduction in the risk of heart attacks and cardiovascular disease is achieved by lowering cholesterol levels and by controlling high blood pressure. Controlling glucose can at best provide only a small further reduction in risk that may take many years to become apparent.


But that does not mean that so-called tight control of blood glucose levels is not a good idea. Data from two of the studies confirm earlier findings that such control reduces the risk of kidney and eye disease, also complications of diabetes.

New data from the VA study also show that severe hypoglycemic events -- exceptionally low levels of blood sugar that lead to blackouts or a change of consciousness -- should be avoided at all costs.

The new findings on hypoglycemia may help to explain the sharp differences between the two earlier studies, experts said.

And finally, data from the three studies may ease concerns about the oral diabetes drug rosiglitazone.


An analysis last year of many previous studies suggested that the drug, sold under the brand name Avandia, increases the risk of heart attacks and death.

But the new studies showed no such effects from the drug, which is still in wide use.

Conflicting findings

Researchers have been looking for ways to reduce the cardiovascular risks of diabetes because 65% to 70% of deaths among Type 2 diabetics are caused by heart problems. Studies in Type 1 diabetics, whose pancreases are no longer able to produce insulin, suggest that tight control of glucose levels reduces risks. But studies in Type 2 diabetics, whose bodies become desensitized to insulin and require ever-larger doses, have been more complicated.


The current controversy began in February when leaders of a government-sponsored study called Action to Control Cardiovascular Risk in Diabetes, or ACCORD, terminated it prematurely after concluding that intensive treatment increased risk of death 22%.

That study of 10,251 people was aimed at getting levels of glycated hemoglobin A1c -- an indirect measure of average glucose levels over the preceding two to three months -- down to about 6%, the same level found in healthy people. This was achieved by using a combination of oral drugs and insulin.

Diabetics in the control arm of the study used the same drugs, but in lower doses, to achieve A1c levels of 7% to 7.5%, which are typical of the levels most diabetics achieve. All of the subjects also took statins to lower cholesterol levels and antihypertensives to reduce blood pressure.

A week later, leaders of a similar Australian study called Action in Diabetes and Vascular Disease, or ADVANCE, said a preliminary analysis of their results showed no increased risk of death.


Their study, which enrolled 11,140 people in Australia, Europe and Asia, had the same overall design as the ACCORD trial, but used a different selection of drugs.

There was “a trend toward a reduction in cardiovascular events” in the intensive control group, but it was not statistically significant, Dr. Stephen MacMahon of the George Institute for International Health in Sydney said at the diabetes association meeting.

Results from the ACCORD and ADVANCE trials were also published online over the weekend by the New England Journal of Medicine, along with several commentaries that attempt to explain the sharp difference in conclusions.

Among the suggested possibilities were the higher incidence of hypoglycemia in the ACCORD study, the use of different drugs, the speed with which blood sugar levels were initially brought down to target levels, and the overall health of the patients.


Many critics thought that it was not the tight control of blood sugar that was the problem, but how it was achieved.

“It’s not the endpoint; it’s how you get there,” said Dr. James Underberg of New York University School of Medicine, who was not involved in either study.

The third study, the Veterans Affairs Diabetes Trial, is the smallest of the three, with about 1,800 patients enrolled. But at 7 1/2 years, it is the longest-running one.

The study, led by Dr. William C. Duckworth of Phoenix VA Health Care System, was similar in design to the others.


The team had expected 650 to 700 cardiovascular events or deaths among all the participants, but only 263 occurred in the standard treatment group and 231 in the intensive group.

“We believe this was largely due to the excellent blood pressure control, lipid control, improved diet and exercise, and treatment with aspirin,” Duckworth said.

An important predictor

The difference between the two groups is not statistically significant, but it appears to be increasing with time, Duckworth said. The team will continue to monitor the participants.


The preliminary analysis of the results found that the most important predictor of a heart attack or death was a severe hypoglycemic event in the previous three months.

“Many people had suspected hypoglycemia was the problem, and we found it,” Duckworth said.

About 95% of the patients had one hypoglycemic event during the study, said statistician Thomas Moritz of the Hines VA Hospital in Illinois.

But those who had more than one such event had more than double the risk of cardiovascular death and triple the risk of any type of death, he said.


The team also found that the time between diagnosis of diabetes and the start of intensive treatment was crucial.

“If we treat patients with intensive care early in the disease, we get a rather large benefit,” Duckworth said. “If we wait until they have had it for 15 years, the effects go away and the statistics show what may be a detrimental effect of glucose control.”

The “simple answer,” he added, “is to just start treating early, as soon as the diagnosis is made. . . . And hypoglycemia should be avoided, no matter what treatment is used.”