Prospective study of dietary fiber intake and breast cancer risk in postmenopausal women: The Multiethnic Cohort Study

B59

There is evidence to suggest that dietary fiber intake may play an important role in estrogen metabolism and therefore may have a beneficial effect on breast cancer incidence. However, epidemiologic studies to date are equivocal. Recent results from our group suggest that particular plant foods such as grapefruit, which are components of total dietary fiber as quantified in nutritional epidemiology studies, may be deleterious in relation to breast cancer and therefore may need to be accounted for when examining dietary fiber in association with cancer risk.
 >This study investigated the hypothesis that dietary fiber intake, total or a specific fraction (e.g., soluble or insoluble non-starch polysaccharides (NSP)) is inversely associated with breast cancer risk. We investigated this hypothesis among postmenopausal women in the Multiethnic Cohort study, a prospective cohort that includes women from five main racial/ethnic groups with a broad range of dietary fiber intake. A total of 1,657 incident breast cancer cases were included.
 >After adjustment for all known risk factors plus total energy, saturated fat, red meat, and grapefruit intakes, intakes of soluble and insoluble NSP were associated with a reduced risk of breast cancer (p_trend = 0.02 and 0.01, respectively). The risk estimate for the highest quintile of insoluble NSP intake was statistically significant (RR = 0.82, 95% CI = 0.67-0.99). However, without adjustment for grapefruit intake, only the trend for insoluble NSP was statistically significant. We examined the associations between breast cancer risk and NSP from particular food sources, e.g., NSP from fruit or NSP from grains. With all adjustments named above, high intake of NSP from fruit + vegetables was significantly associated with a reduced risk of breast cancer (RR=0.82; 95% C.I. 0.67-1.00; p_trend = 0.01). No other significant associations were observed.
 >We examined whether the association of total NSP intake and breast cancer risk was modified by body mass index (BMI), hormone therapy (HT) use, family history of breast cancer, and alcohol use. Because there was no evidence of association with low to moderate consumption of NSP, we combined quintiles 1-3 for this stratified analysis. Reduced risk of breast cancer associated with high intake of total NSP was stronger among current HT users (RR for the highest quintile=0.70; 95% CI: 0.55, 0.89) than among never or former HT users (RR=0.80; 95% CI: 0.64, 0.99). The reduced risk associated with high intake of total NSP was also stronger among women without a family history of breast cancer (RR=0.76; 95% CI: 0.64, 0.91) than among women with a family history (RR for the highest quintile=0.85; 95% CI: 0.56, 1.28). Although the inverse associations appeared stronger among current HT users and among women without a family history, the tests for interaction for both factors were not significant. There was no statistically significant interaction between total NSP intake and BMI or alcohol use.
 >This prospective study provides evidence of a protective association between high dietary fiber intake and breast cancer risk in postmenopausal women. Our results suggest that specific plant foods which may be deleterious in relation to breast cancer risk need to be adjusted for in the statistical model to provide a more accurate estimate of risk associated with total dietary fiber intake.