Multivariate Analysis of Risk Factors for Coronary Heart Disease

Abstract

For a prospective study of risk factors for coronary heart disease (CHD), defined as acute nonfatal myocardial infarction and acute fatal CHD, 973 men, all aged 50 years, were recruited from a general Swedish urban population. Of the 855 who agreed to participate, 834 who showed no signs of CHD were selected and observed for nine years and four months. Autopsies were performed on all except two of the 55 patients who died during the study. Forty-four men developed clinical manifestations of CHD during this time; 19 of them died. Using a multiple logistic model, we analyzed nine probable risk factors. The presence of high serum cholesterol, smoking, high systolic blood pressure, dyspnea, and registration by the Temperance Board increased the risk of events related to CHD significantly. The presence of high serum triglycerides, high hematocrit readings, and increased geographic mobility did not cause a significant increase in events related to CHD. Those with sedentary jobs showed a slightly increased tendency to develop manifestations of the disease. The predictive power of the logistic function (with cholesterol, smoking, and systolic blood pressure) was tested in another randomly selected population sample of 5,146 men aged 51-55 years and found to be very accurate. The 10% (decile) of the population that had the highest risk of clinical manifestations of coronary disease and the 10% who had the lowest risk were defined. Manifestations of coronary disease occurred 29 times more frequently in the highest decile than in the lowest decile.