We reviewed epidemiologic evidence on the relationship between four an
tioxidant micronutrients (vitamin A, vitamin C, vitamin E, and seleniu
m) and breast cancer risk. Available data support a modest protective
effect of vitamin A, although more studies are needed to examine furth
er this association and to assess the relative contributions of prefor
med vitamin A (retinol) and carotenoids. In addition, the possibility
that some other component of vitamin A-rich foods may account for this
observed association should be explored. Data on the relationship bet
ween vitamins C and E and breast cancer risk are limited and inconsist
ent, and further information is necessary. A substantial body of evide
nce indicates a lack of any appreciable effect of selenium intake on b
reast cancer risk, at least within the range of human diets. Future ob
servational studies should ideally be prospective in design, as prospe
ctive studies are less prone to selection and recall bias than are cas
e-control studies, and should address methodologic issues such as conf
ounding by other micronutrients and appropriate storage conditions of
blood specimens. Although hypotheses relating micronutrient intake to
risk of breast cancer should be tested in randomized trials, ethical a
nd logistical constraints make these studies difficult to perform.