Abstract 640: BMI as Surrogate for Cholesterol to Calculate Coronary Risk and to Decide on Lipid Lowering Medication: Improved Results with Carotid Total Plaque Area.
Background: The Framingham risk calculator (FRAM) offers the possibility to use the body-mass-index (FRAMb) instead of total and HDL cholesterol (FRAMc). TPA, total plaque area of carotid arteries, may additionally serve to correct for inaccuracies due to unknown cholesterol measurements (FRAMb-TPA;FRAMc-TPA).
Material and Methods: Participants were recruited from a free checkup program offered by the Swiss Vascular Risk Foundation and included healthy subjects aged > 45 years from the ongoing Cordicare II Study. Predictive risk was compared using the FRAM, FRAMb-TPA and FRAMc-TPA using linear regression models, kappa statistics and areas under the curve analysis (ROC).
Results: Of 1,000 participants, 47% were females, mean age 60±9 years. 3.3% had Diabetes Mellitus type II, 11% were smokers, and 18% had family history of premature coronary disease. Mean systolic blood pressure was 134±15 mmHg and total cholesterol, HDL-cholesterol, LDL-cholesterol and triglycerides were 5.7±1.1, 1.6±0.4, and 3.5±0.9 and 1.4±0.7 mmol/l respectively. Mean body mass index was 26±4 kg/m2 and total plaque area was 51±51 mm2. Linear regression between FRAMc and FRAMb ten-year coronary risk showed an R2 = 0.89 (p<0.0001) and a Kappa coefficient of 0.72 (p<0.0001). This correlation was further enhanced when comparing FRAMc-TPA to FRAMb-TPA (R2 = 0.94, p<0.0001, wKappa 0.85, p 20% was assessed by ROC analysis and showed an area under the curve (AUC) of 0.98 (95% CI = 0.98 - 0.99, p<0.0001). FRAMb-TPA showing a risk of > 10% had a sensitivity and specificity of 92% and 56% (accuracy 75%) with the highest Youden’s Index, respectively, for the indication to lower LDL cholesterol according to the NCEP III guidelines and a risk assessment defined by FRAMc+TPA.
Conclusion: When using the FRAM coronary risk function that uses BMI instead of total and HDL cholesterol and results from carotid plaque imaging, we observed a very high correlation, agreement and accuracy of this new method. Within this diagnostic setting, cholesterol profiles can be replaced by BMI without a relevant loss in coronary risk stratification. The indication for a lipid lowering medication defined by FRAMb-TPA is highly sensitive and moderately specific for a coronary risk cutoff of 10% or more.
- © 2012 by American Heart Association, Inc.