[Risk factors for breast cancer and their use in early detection. Results of a case control study].

Doubtful and established breast cancer risk factors were checked in a case/control study. Multivariate analysis with the logistical regression approach resulted in a usable separation of the cases from the controls on the basis of the five dichotimized variables: age at birth of first child, family history, Quetelet index, breast complaints, and a behavior score. On the basis of probability, the risk patterns were summarized in six risk categories, and a maximum 70-fold risk was thus established. The proportion of incidence represented by the risk groups was then estimated from distribution in the control group. Approx. 5% of new cases originate in 30% of the female population aged under 30, but approx. 51% originate in a mere 10%. Hence it is logical to use risk factors for early detection based on risk.