Veterans who received initial diabetes treatment with metformin experienced lower rates of death and major cardiovascular events than those who received sulfonylurea, but it is not clear whether sulfonylurea raises risk or metformin is protective. Those are the findings of a large, retrospective cohort study published in the November issue of Annals of Internal Medicine.
Many patients with diabetes die of cardiovascular disease (CVD), but some commonly prescribed diabetes medications, in particular sulfonylurea, have been associated with increased risk of CVD. Christianne L. Roumie, MD, of the Veterans Health Administration (VHA) in Nashville, Tenn., and colleagues, sought to explicate the relative risks of CVD associated with metformin and sulfonylurea use in diabetic patients.
The researchers used data from the VHA to define a cohort of veterans 18 years or older who began diabetes therapy between Oct. 1, 2001, and Sept. 30, 2008, and received care from the VHA. The researchers defined care as one encounter or prescription fill every 180 days for at least the past year. They excluded patients with other serious medical conditions, cocaine use or baseline creatinine levels of 1.5mg/dL or greater.
The final cohort comprised 269,921 patients, 97 percent were men and 75 percent were white. Seventy-third percent had no history of CVD at the time of their initial prescription for diabetes drugs. The study analyzed the results of only those patients with initial prescription for metformin (50 percent) or sulfonylurea (40 percent) and excluded those patients receiving combination metformin/sulfonylurea or other diabetes medications.
The median age was 62 among the metformin users and 67 years among sulfonylurea users. The characteristics of the two groups were similar after propensity score matching.
The study followed the patients from first prescription until a change in anti-diabetes medication, the 90th day without drugs in hand, an outcome or a censoring event (censoring events included high serum creatinine levels, lack of contact with VHA for more than 180 days and the end of the study), whichever came first.
The primary outcome was a composite of hospitalization for acute MI or stroke, or death. The researchers used the VHA Vital Status file to determine mortality. Secondary endpoints were a composite of acute MI and stroke, with death as a censoring event rather than an endpoint.
The primary outcome rate was 18.2 per 1,000 person-years in the sulfonylurea group and 10.4 per 1,000 person-years for the metformin group. Rates of CVD events excluding death were 13.5 per 1,000 person-years for the sulfonylurea group and 8.2 per 1,000 person-years for the metformin group. "Using adjusted rate differences, we estimated 2.2 more CVD events and deaths and 1.2 more CVD events per 1000 person-years of sulfonylurea use compared with metaformin use," the authors wrote.
They noted that confounding by indication could be possible, if patients with CVD risk characteristics were more likely to use one drug or the other. However, the authors asserted that there would need to be a very large imbalance between the two groups to influence the results of their analysis. The authors also pointed out the study population was overwhelmingly white and male, and cautioned against applying these results to the population as a whole.
In addition, the authors conceded that the study results do not clarify whether sulfonylurea causes harm or metformin is protective. Even so, "these observations support the use of metformin for first-line diabetes therapy and strengthen the evidence about the cardiovascular advantages of metformin compared with sulfonylurea," they wrote.