Abstract
BACKGROUND: On the basis of hypothesized protective effect, we examined the effect of soy foods on estrogens in nipple aspirate fluid (NAF) and serum, possible indicators of breast cancer risk. METHODS: In a crossover design, we randomized 96 women who produced 10 μL or more NAF to a high- or low-soy diet for 6 months. During the high-soy diet, participants consumed 2 soy servings of soy milk, tofu, or soy nuts (∼50 mg of isoflavones per day); during the low-soy diet, they maintained their usual diet. Six NAF samples were obtained using a FirstCyte aspirator. Estradiol (E(2)) and estrone sulfate (E(1)S) were assessed in NAF and estrone (E(1)) in serum only, using highly sensitive radioimmunoassays. Mixed-effects regression models accounting for repeated measures and left-censoring limits were applied. RESULTS: Mean E(2) and E(1)S were lower during the high-soy than the low-soy diet (113 vs. 313 pg/mL and 46 vs. 68 ng/mL, respectively) without reaching significance (P = 0.07); the interaction between group and diet was not significant. There was no effect of the soy treatment on serum levels of E(2) (P = 0.76), E(1) (P = 0.86), or E(1)S (P = 0.56). Within individuals, NAF and serum levels of E(2) (r(s) = 0.37; P < 0.001) but not of E(1)S (r(s) = 0.004; P = 0.97) were correlated. E(2) and E(1)S in NAF and serum were strongly associated (r(s) = 0.78 and r(s) = 0.48; P < 0.001). CONCLUSION: Soy foods in amounts consumed by Asians did not significantly modify estrogen levels in NAF and serum. IMPACT: The trend toward lower estrogen levels in NAF during the high-soy diet counters concerns about adverse effects of soy foods on breast cancer risk.