[The prediction of coronary heart disease. A mathimatical model applied to Italian field studies (author's transl)]

The multiple logistic function (MLF) was applied for the estimation of coronary heart disease (CHD) - "hard criteria" in 5 years as a function of several risk factors in two rural Italian areas belonging to the "Seven Countries Study". CHD was estimated with solutions including 4, 5 and 11 factors respectively. Only serum cholesterol showed a statistically significant contribution; but other factors, like age and physical activity, very likely contributed to a better discrimination between "cases" and "non-cases". By increasing the number of factors the discrimination tends to improve anyway. In the solution with 11 factors, 46% of the observed cases was found in the upper 10% of the estimated risk probabilities; 57% in the upper 20%; while no cases were found in the lowest 20% of the same rank probability list. For practical purposes of selecting high risk individuals for preventive action, a simple 5 factor solution, including age, systolic blood pressure, serum cholesterol, consumption of cigarettes and physical activity is suggested.