Third, researchers have focused on so-called gene-environment interactions, reasoning that some exposures that have relatively weak associations with breast cancer in the general population may have a much stronger effect in subgroups with a specific genetic make-up.