Study of Atherosclerosis with Ramipril and Rosiglitazone - STARR


The goal of the trial was to evaluate treatment with ramipril compared with placebo, and rosiglitazone compared with placebo (in a 2 x 2 factorial design) among patients with impaired glucose tolerance and/or impaired fasting glucose.


Ramipril and rosiglitazone will be more effective in reducing carotid intima-media thickness (CIMT) progression.

Study Design

  • Placebo Controlled
  • Randomized
  • Blinded
  • Parallel

Patients Enrolled: 1,425
Mean Follow Up: Median 3 years
Mean Patient Age: 54 years
Female: 58%

Patient Populations:

  • Patients at least 35 years of age with impaired fasting glucose (110-126 mg/dl) or impaired glucose tolerance (140-200 mg/dl 2 hours after an oral glucose load)
  • Adequate baseline carotid ultrasound


  • Diabetes
  • Cardiovascular disease
  • Intolerance to any of the study medications

Primary Endpoints:

  • Annualized change in the aggregate CIMT

Secondary Endpoints:

  • Annualized change in the common carotid far wall CIMT

Drug/Procedures Used:

Patients with impaired glucose tolerance or impaired fasting glucose were randomized to ramipril (5 mg daily, titrated to 15 mg daily at 1 year; n = 715) versus placebo (n = 710) then to rosiglitazone (4 mg daily, titrated to 8 mg daily at 2 months; n = 709) versus placebo (n = 716).

Concomitant Medications:

At baseline, the use of aspirin was 12%, statin was 9%, and beta-blocker was 16%.

Principal Findings:

Overall, 1,425 patients were randomized. There was no difference in baseline characteristics between the groups. The mean age was 54 years, 58% were women, and body mass index was 30 kg/m2.

At baseline, the aggregate maximum CIMT was 0.75 mm in the ramipril group, 0.76 mm in the ramipril placebo group, 0.75 mm in the rosiglitazone group, and 0.76 mm in the rosiglitazone placebo group. The annualized change in the aggregate maximum CIMT was 0.0083 mm/year in the ramipril group versus 0.0069 mm/year in the ramipril placebo group (p = 0.37) and 0.0063 mm/year in the rosiglitazone group versus 0.0090 mm/year in the rosiglitazone placebo group (p = 0.08).

The annualized change in the common carotid far wall CIMT was 0.0027 mm/year for ramipril versus 0.0047 for the ramipril placebo group (p = 0.26) and 0.0015 mm/year for rosiglitazone versus 0.0058 mm/year for rosiglitazone placebo (p = 0.01).

Composite cardiovascular outcomes occurred in 1.7% of the ramipril group, 2.1% of the ramipril placebo group (p = 0.53), 2.3% of the rosiglitazone group, and 1.5% of the rosiglitazone placebo group (p = 0.34).


Among patients with impaired fasting glucose or impaired glucose tolerance, ramipril did not change the primary outcome, annualized aggregate maximum CIMT, although rosiglitazone was marginally beneficial (p = 0.08). For the secondary outcome, common carotid far wall CIMT, ramipril was again neutral, while rosiglitazone was beneficial (p = 0.01). There was no obvious difference on clinical outcomes.

The use of rosiglitazone in high-risk populations may deserve further study, specifically examining clinical outcomes.


Lonn EM, Gerstein HC, Sheridan P, et al. Effect of ramipril and of rosiglitazone on carotid intima-media thickness in people with impaired glucose tolerance or impaired fasting glucose: STARR (Study of Atherosclerosis with Ramipril and Rosiglitazone). J Am Coll Cardiol 2009;53:2028-35.

Clinical Topics: Noninvasive Imaging, Prevention, Echocardiography/Ultrasound, Diet

Keywords: Prediabetic State, Glucose Intolerance, Body Mass Index, Carotid Intima-Media Thickness, Hypoglycemic Agents, Ramipril, Fasting, Thiazolidinediones

< Back to Listings