ACCORD: Insulin Dose Not Responsible for Increased CV Mortality Risk

June 26, 2013

CHICAGO, IL — The dose of insulin used in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial was not responsible for the increased risk of cardiovascular mortality among patients in the intensive-glycemic-control arm, according to a post hoc analysis of the data [1]. Adjusting for multiple patient and on-treatment parameters, investigators report that the risk of cardiovascular mortality was not significantly elevated among those patients who received higher doses of insulin.

"Our conclusion is basically that given the patient population in ACCORD, and based on how we performed the analysis, we did not confirm insulin as an independent risk factor for cardiovascular death," lead investigator Dr Elias Siraj (Temple University School of Medicine, Philadelphia, PA) told heartwire .

Presenting the results of the analysis here at the American Diabetes Association (ADA) 2013 Scientific Sessions , Siraj said it is difficult to make firm clinical recommendations based on the data.

"The problem with advice to clinicians is that it's very difficult to take one conclusion from one study and automatically translate it to clinical practice," he said. "Number one, remember, this is a post hoc analysis. It's not designed to specifically answer this question. The results contribute to the literature, to the debate, but I can't say this is the conclusive study, although it is an important one."

ACCORD Stopped Because of Mortality Risks

Since ACCORD was stopped in February 2008 because of higher rates of overall and cardiovascular mortality in type 2 diabetics treated with intensive glycemic control, researchers have been attempting to understand the reason for the increased risk. Other investigators have questioned whether hypoglycemia, weight gain, specific medication use, a rapid decline in HbA1c levels in the first year of treatment, or low HbA1c levels per se were responsible for the increased risk, but so far none of these variables have been shown to be associated with mortality or cardiovascular mortality.

To heartwire , Siraj said the group studied the potential role of insulin dose on the risk of death because epidemiological studies have shown that hyperinsulinemia is associated with adverse cardiovascular outcomes. In addition, there are some studies, although they might not be the most rigorous or definitive, showing that insulin treatment might increase the risk of death. These results, however, remain controversial and limited in their interpretation, said Siraj. In addition, patients in the ACCORD trial treated with intensive glycemic control were receiving higher doses of insulin than those in the standard-treatment arm.

In their analysis of ACCORD, the unadjusted data showed that each 1-unit/kg increase in insulin dose was associated with an 83% higher risk of cardiovascular mortality. These risks were higher for patients receiving basal insulin and higher still for those receiving a bolus of insulin. However, after adjustment for multiple baseline covariates, the risk was attenuated and no longer statistically significant. Further adjustment for other drug treatment, weight change, average HbA1c, and glycemic treatment arm attenuated the association further.

"Once we adjusted, the association disappeared," said Siraj.

During the ADA presentation, the issue was raised that the model used to adjust for covariates might not be the correct one and that the association between insulin dose and cardiovascular mortality might be lost in such "overparameterization." However, Siraj said the group felt they needed to address the baseline characteristics of the treated patients. They identified 14 different baseline variables that were statistically relevant, and when these were added, the association was no longer statistically significant. Adding in on-treatment factors to the model did not change the results.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: