In an analysis of the study Brind said the researchers probably underestimated the breast cancer risk associated with abortion because of ???selection bias,??? in that a disproportionate number of women with modern lifestyles, including abortions, were likely overrepresented among the controls.