Application of regression analysis and classification trees in calculating additional population risk of ischemic heart disease