Fig. (2). Interaction between probiotic beverage and soy isoflavone consumption.

* Calculated using conditional logistic regression. 306 cases and 661 controls were matched for residential area and adjusted for age (40s and 50s), educational background, physical activity level, benign mammary tumor, family breast cancer history, menarche, number of births, breastfeeding, use of female hormone, birth weight, Body Mass Index around the age of 20, smoking and energy intake. One control was excluded due to lack of adjustment factor.