To characterize bioavailability of soybean isoflavones, proposed anticarcin
ogenic food components, eight women, ages 20-41 y, were fed 0.9 mg isoflavo
nes/kg body wt from soymilk at 0730, 1230 and 1730 h for 1 d, Subjects cons
umed three background diets in random order: a diet prepared for them (basi
c foods diet) or a self-selected diet at the specified times, or a self-sel
ected diet eaten ad libitum, In a second study, women were fed single isofl
avone doses of 0.8-1.4 mg/kg in breakfast casseroles containing tofu, tempe
h, cooked soybeans or texturized vegetable protein. Both studies were condu
cted in randomized, cross-over designs. Plasma, urine and fecal isoflavones
were measured by reverse-phase HPLC, After consumption of background diets
, 48-h urinary recovery of daidzein (D) was 26-27%, and of genistein (G), 1
8-20% of the dose given with each diet. At 24 h after consumption of differ
ent background diets, plasma D and G concentrations were similar (1.4 +/- 0
.7 mmol/L) and were not affected by diet selection. Urinary recoveries of D
over 24 h from the various soy foods were 38-51%, and of G, 9-16% of the d
ose given. In both studies, urinary recovery of D was significantly greater
than that of G. Only a few percentage of the total isoflavone dose was rec
overed in feces, probably due to bacterial breakdown of these compounds. Th
erefore, isoflavone bioavailability may not be affected by choice of backgr
ound diet or food source of isoflavones.