Obesity and risk of myocardial infarction in a sample of European males. Waist to-hip-ratio presents information bias of the real risk of abdominal obesity

Background: Obesity is a coronary risk factor associated to myocardial infarction although waist to-hip-ratio has shown higher predictive power. Objective: The aim of this study was a Receiver Operating Characteristic anthropometric analysis in infarcted males to identify the strength of association for simple measurements, obesity and indicators such as, waist to-hip-ratios, waist to-height-ratios and conicity index. Methods: Case-control study of myocardial infarction in European males. One hundred and twelve cases and 112 controls aged 30-74 years were enrolled. We measured weight, height, waist circumference, umbilical waist circumference and hip circumference. We calculated various anthropometric indicators. We obtained the areas under the ROC curves, the odds ratio and correlations for measurements and anthropometric indicators. Results: Body mass index [AUC: 0.686, 95% Cl (0.616-0.755); OR: 3.3], waist circumference [AUC: 0.734, 95% Cl (0.668-0.800); OR: 5.7], height [AUC: 0.623, 95% Cl (0.550-0.696); OR: 2.3], hip circumference [AUC: 0.555, 95% Cl (0.479-0.631); OR: 1], waist to-hip-ratio [AUC: 0.796, 95% Cl (0.737-0.855); OR: 9.9], umbilical waist to-hip-ratio [AUC: 0.830, 95% Cl (0.729-0.847); OR: 5.5], umbilical waist to-height-ratio [AUC: 0.788, 95% Cl (0.729-0.847); OR: 7.5], conicity index [AUC: 0.795; 95% Cl (0.738-0.853); OR: 9]. The correlations for waist to-height-ratios and conicity index were strong (all r >= 0.85; p < 0.001). Conclusions: Waist and height are measurements of associated independent risk. Hip circumference does no show discriminatory power. Obesity and waist-ratios are associated to myocardial infarction with different strength. Between other indicators, general obesity is more weakly associated. Waist to-hip-ratios present the best ROC curves but it occur information bias of their predictive power of risk. Umbilical waist to-height-ratio and conicity index present high discriminatory power and the best anthropometric risk correlations that support its use for the identification of obesity as risk factor associated to myocardial infarction and in all strategies for coronary health promotion.