« Back

Study identifies potential predictors of diabetes among patients with high CVD risk

Wednesday, November 12 2008 | Comments

---

What's This? By Patrice La Vigne

Cardiovascular disease (CVD) and diabetes mellitus often occur concurrently, and while administering an oral glucose tolerance test (OGTT) to all patients with a high CVD risk would be ideal for identifying undiagnosed diabetes, Dr. Mary Angelyn Bethel, lead study author, told VerusMed that there may be other tools available to help determine the best candidates for an OGTT.

The investigators noted that the OGTT is too expensive and inconvenient to recommend for all high-risk patients.

Using the patient population from the NAVIGATOR study, Dr. Bethel and colleagues assessed the prevalence of undiagnosed diabetes among 42,149 outpatients with high cardiovascular risk. The patients were aged 55 years or older with >=2 cardiovascular risk factors or were aged 50 years or older with known CVD.

Of the 35,744 patients with valid OGTT results, 22% had diabetes (defined as fasting plasma glucose [FPG] >=126 mg/dL and 2-hour glucose values >=200 mg/dL), 20% had normal glucose tolerance, 28% had impaired glucose tolerance, and 30% had impaired fasting glucose. Dr. Bethel told VerusMed that these results were similar to results found among other study populations with high CVD risk.

The investigators then created 2 models to identify predictors of undiagnosed diabetes among these patients. The basic model used patient history variables determined during a routine physical (eg, age, sex, body mass index, blood pressure [BP], and presence of CVD), while the enhanced model added routine laboratory tests commonly available in a cardiologist's office (eg, low density lipoprotein [LDL] cholesterol, triglycerides, and FPG).

In both models, age, systolic BP, and known CVD were statistically significant predictors of undiagnosed diabetes. In the enhanced model, other laboratory tests also proved to be statistically significant in predicting undiagnosed diabetes, such as FPG (odds ratio [OR], 4.31; 95% CI, 3.76-4.93), triglycerides (OR, 1.32; 95% CI, 1.24-1.4), and LDL cholesterol (OR, 0.85; 95% CI, 0.8-0.9).

Dr. Bethel told VerusMed that the area under the curve for the enhanced model was 0.75, which signified that OGTT was a good test, although she said it is probably not sufficient for clinical use.

"Although it's a pretty good test--you're going to be right most of the time--you're still going to be missing a lot of people with undiagnosed diabetes," she told VerusMed. "You're going to miss the chance to intensify risk factor control for CVD and also obviously miss the chance to treat diabetes and prevent microvascular disease."

She added, "mostly we're just asking the people to think," and to get "diabetes on the radar screen" for these patient populations. (Abstract Poster Session 1182.)
 

Print  |  E-mail