Predictors of arteriographically defined coronary stenosis in the Honolulu Heart Program. Comparisons of cohort and arteriography series analyses.

The purpose of this study was to determine if the major risk factors for clinical myocardial infarction also predicted coronary artery stenosis as defined by arteriography. Of a cohort of 7,591 men who were free of cardiovascular disease at entry, 357 had arteriographic studies during a 20-year follow-up period. Risk factor levels were therefore known prior to the onset of clinical symptoms and arteriographic studies. Men with arteriograms were divided into groups with and without prior clinical myocardial infarction. High blood pressure, serum cholesterol, obesity, and low alcohol intake predicted both severe coronary stenosis and incident myocardial infarction, thus indicating that these variables were associated with clinical events through the underlying process of atherosclerosis. Dietary intake of cholesterol and serum glucose also had similar but not always statistically significant patterns of association with both coronary stenosis and myocardial infarction. In contrast, serum triglyceride and cigarette smoking predicted clinical myocardial infarction, but not severe coronary stenosis. This suggests that these variables play a stronger role in the precipitation of acute clinical events than in the underlying process of atherosclerosis. The findings were quite different for several risk factors when analyzed in a case-control format using the arteriography series from this same data set. Examination of possible explanations for the differences raises questions concerning the use of arteriography series for etiologic studies of coronary atherosclerosis.