However, other experts question the study's conclusions and claim it is at odds with other better-designed studies that show the combination of metformin and insulin is both safe and effective.
"Insulin remains a reasonable option for patients who have very high glucose [blood sugar] or who desire flexible and fast blood sugar control, but most patients taking metformin prefer to delay starting insulin," said lead researcher Dr. Christianne Roumie, an associate professor of internal medicine and pediatrics at Vanderbilt University in Nashville, Tenn.
"The current study suggests that adding a sulfonylurea to metformin should be preferred to adding insulin for most patients who need a second diabetes drug," she said. Sulfonylureas include glibenclamide (Micronase), glimepiride (Amaryl), glipizide (Glucotrol) and others. They work by stimulating the body to make more insulin.
Roumie's team found that, compared with those who added a sulfonylurea, those who added insulin to metformin had 30 percent higher odds of heart attack, stroke and death from any cause during the study period. "Although new heart attacks and strokes occurred at similar rates in both groups, mortality was higher in patients who added insulin," she said.
Roumie said she doesn't know why the rate of deaths was higher among study patients taking insulin. "We have a number of studies planned to examine possible mechanisms. We are investigating type 2 diabetes outcomes associated with blood sugar swings and with episodes of hypoglycemia (low blood sugar) tied to insulin," she said.
Sulfonylureas can also cause low blood sugar levels, according to the American Diabetes Association.
The report was published June 11 in the Journal of the American Medical Association.
For the study, Roumie's team compared the outcomes among people receiving metformin and insulin with those taking metformin and sulfonylureas.
Using U.S. Veterans Health Administration (VA), Medicare, and National Death Index data, they compared the risk of heart attack, stroke, or all-cause death between the different therapies. Among more than 178,000 people taking metformin, almost 3,000 added insulin and nearly 40,000 added a sulfonylurea.
Focusing on about 2,400 patients who added insulin to metformin and just over 12,000 patients who added a sulfonylurea, Roumie's group found that during an average of 14 months of follow-up, the rate of heart attacks and strokes was similar in both groups. However, the rate of death from any cause was higher among those patients taking insulin, the investigators reported.
Other experts dispute the study's conclusions.
Dr. Robert Ratner, chief scientific and medical officer at the American Diabetes Association, said, "I disagree with metformin and insulin being a bad combination."
Ratner said that there are other factors that could explain the results of this study. Those placed on insulin had higher blood sugar levels and had other serious medical conditions, he said.
"These were not equivalent populations. If you're sicker, then it's not surprising you're going to have worse outcomes," Ratner said.
Ratner noted that a randomized trial that compared these same drug combinations, called the ORIGIN trial (Outcome Reduction With Initial Glargine Intervention), which followed some 12,000 patients over seven years, came to a different conclusion.
"What they found in ORIGIN was there was no difference in heart attacks or strokes between the two groups, there was no difference in cancer, and there was no difference in all-cause deaths," he explained. "So the better study showed no difference."
Ratner said, based on this current study, there is no reason to change treatment.
Dr. Joel Zonszein, director of the Clinical Diabetes Center at Montefiore Medical Center in New York City, agreed. "You cannot make any conclusions from this study," he said.
"We are using much newer medications," Zonszein said. "In the VA, they don't use those new medications. They use medications from the last century. The VA continues to use inexpensive therapy that is not very effective."
Zonszein said that new drugs such as Victoza or Januvia are very expensive and often are not covered by insurance, or if they are, they have very high co-pays.
Ratner added, "We really don't know what the best drug is to add after metformin." The American Diabetes Association recommends an individualized patient approach that takes into account the patient's condition as well as the cost of treatment, he said.
A new trial is getting underway that might better pinpoint the most effective and safe treatment, Ratner said. The new study will compare adding insulin or a sulfonylurea or newer drugs like Victoza or Januvia to metformin. "Hopefully, we will have a much better answer in the next several years," Ratner added.
For more about type 2 diabetes medications, visit the American Diabetes Association.