39898

This paper considers the problem of searching for dependences between the cancer occurrence and associated diseases. A statistical method that yields blocks of diseases with the most significant influence on cancer morbidity is described. Based on American data on causespecific mortality, we select the diseases that have the maximum differences between the distributions of associated diseases among people who died from cancer or have cancer as an associated disease and among people who did not have cancer. A medical interpretation of the obtained results is discussed.