Published June 28, 2013 at Healio Cardiology
CHICAGO — Results from a post-hoc analysis suggest that higher insulin doses were not responsible for the increased CV mortality found in the ACCORD trial, a presenter said at the ADA Scientific Sessions.
Main results from ACCORD demonstrated an association between increased all-cause and CV mortality in patients assigned to intensive treatment for diabetes, but several post-hoc analyses have been unable to pinpoint the cause, according to Elias S. Siraj, MD, FACP, FACE, associate professor of medicine, director of the diabetes program and director of clinical endocrinology at Temple University School of Medicine.
Elias S. Siraj
“This presented the diabetes community with a huge puzzle that challenged us to figure out why we are seeing these results,” Siraj, who presented the data, said during a presentation.
Researchers have, however, identified a link between increased mortality and higher HbA1c in the intensive treatment arm, prompting Siraj and colleagues to hypothesize that higher doses of insulin in these patients may have contributed to increased CV mortality.
The researchers examined insulin exposure data in units/kg of body weight from 10,163 patients with a mean follow-up of 5 years. Data from an unadjusted univariate analysis implicated insulin exposure in the increased risk for CV mortality. HRs for all insulin, basal insulin and bolus insulin were 1.83 (95% CI, 1.45-2.31), 2.29 (95% CI, 1.62-3.23) and 3.36 (95% CI, 2.00-5.66), respectively. However, after adjustment for 14 baseline characteristics, including age, history of CVD, complications of diabetes and baseline HbA1c, these associations were no longer statistically significant. HRs declined to 1.21 (95% CI, 0.92-1.6) for all insulin, 1.3 (95% CI, 0.87-1.94) for basal insulin and 1.65 (95% CI, 0.88-3.11) for bolus insulin, according to the study abstract.
Siraj reported that, consistent with prior analyses of all-cause mortality, higher HbA1c remained associated with higher CV mortality before (HR=1.38;. P<.0001) and after adjustment for baseline variables and insulin exposure (HR=1.49; P<.0001).
“In conclusion, after adjustment for covariates, insulin dose was not associated with increased CV morality, and these results do not support the hypothesis that higher dose of insulin is an independent risk factor for CV mortality,” Siraj said. – by Melissa Foster
For more information:
Siraj ES. #386-OR. Presented at: ADA Scientific Sessions; June 21-25, 2013; Chicago.
Disclosure: Siraj reports receiving honoraria for speaking or consulting from Boehringer Ingelheim, Merck and Sanofi.