A comparison between BMI and conicity index on predicting coronary heart disease: The Framingham Heart Study

PURPOSE: This study examined the relationship of mortality and morbidity of coronary heart disease with body mass index (BMI) and Conicity index (CI). METHODS: Among 5209 Framingham Heart Study participants, 1882 men and 2373 women had waist and weight measurement at the 4(th) examination period and height measured on the 5(th) visit. These were used for BMI and CI. RESULTS: During a 24-year follow-up, 597 men and 468 women developed CHD an d 248 men and 150 women died from CHD associated causes. In men the relativ e risks (RR) (95% confidence interval) adjusted for age, hypertension, diab etes, smoking status, and total cholesterol for CHD incidence in 2(nd), 3(r d), and 4(th) quartiles of BMI were 1.28 (1.0, 1.65), 1.45 (1.13, 1.86), an d 1.53 (1.19, 1.96). The RR for CHD incidence in the 4(th) quartile of BMI in women was 1.56 (1.16, 2.08). No CI quartiles were risk factors for CHD i ncidence. There was 86% higher risk of CHD related death in the 4(th) quart ile of BMI than the 1(st) quartile of BMI in women. In men no significantly higher risks of death were found across the quartiles of BMI. No associati ons were found between CI quartiles and CHD mortality. CONCLUSIONS: Obesity as measured by BMI is an important risk factor for CHD incidence in men and women and for CHD mortality in women. CI was not asso ciated with an increase in CHD incidence or mortality. Thus, BMI is a bette r marker than CI for predicting CHD incidence and mortality. Ann Epidemiol 2000;10:424-431. (C) 2000 Elsevier Science Inc. All rights reserved.