Plasma folate, vitamin B6, vitamin B12, homocysteine, and risk of breast cancer.

BACKGROUND In several epidemiologic investigations, folate intake has appeared to reduce the elevated risk of breast cancer associated with moderate alcohol consumption. However, data relating plasma folate levels to breast cancer risk are sparse. We investigated the association between plasma folate and other vitamins with breast cancer in a prospective, nested case-control study. METHODS Blood samples were obtained during 1989 and 1990 from 32 826 women in the Nurses' Health Study who were followed through 1996 for the development of breast cancer. We identified 712 breast cancer case patients and selected 712 individually matched control subjects. Dietary information was obtained using food frequency questionnaires given in 1980, 1984, 1986, and 1990. Logistic regression was used to estimate the relative risks (RRs) of breast cancer (after adjustment for potential risk factors), and a generalized linear model was used to calculate the Pearson correlation coefficients between plasma estimates of folate, vitamin B(6), vitamin B(12), and homocysteine, and intakes of folate, vitamin B(6), and vitamin B(12). All statistical tests were two-sided. RESULTS The multivariable RR comparing women in the highest quintile of plasma folate with those in the lowest was 0.73 (95% confidence interval [CI] = 0.50 to 1.07; P(trend) =.06). The inverse association between plasma folate and breast cancer risk was highly statistically significant among women consuming at least 15 g/day (i.e., approximately 1 drink/day) of alcohol (multivariable RR = 0.11, 95% CI = 0.02 to 0.59 for highest versus lowest quintile) in contrast with that of women consuming less than 15 g/day (multivariable RR = 0.72, 95% CI = 0.49 to 1.05). The multivariable RR comparing women in the highest quintile of plasma vitamin B(6) levels with those in the lowest quintile was 0.70 (95% CI = 0.48 to 1.02; P(trend) =.09). Plasma vitamin B(12) levels were inversely associated with breast cancer risk among premenopausal women (multivariable RR = 0.36, 95% CI = 0.15 to 0.86 for highest versus lowest quintile) but not among postmenopausal women. Plasma homocysteine was not associated with breast cancer risk. CONCLUSIONS Higher plasma levels of folate and possibly vitamin B(6) may reduce the risk of developing breast cancer. Achieving adequate circulating levels of folate may be particularly important for women at higher risk of developing breast cancer because of higher alcohol consumption.