Dietary cadmium and risk of invasive postmenopausal breast cancer in the VITAL cohort.

PURPOSE: The aim of the study was to estimate the association between dietary intake of cadmium, a carcinogenic heavy metal, and risk of invasive postmenopausal breast cancer. METHODS: Study subjects were 30,543 postmenopausal women in the VITamins And Lifestyle (VITAL) cohort who completed a food frequency questionnaire (FFQ) at baseline (2000-2002). Dietary cadmium consumption was estimated by combining FFQ responses with US Food and Drug Administration data on food cadmium content. Incidence of invasive breast cancer was ascertained through linkage of the cohort to the western Washington Surveillance, Epidemiology, and End Results cancer registry through 31 December 2009. Cox regression was applied to estimate adjusted hazard ratios (aHRs) and 95 % confidence intervals (CIs) for breast cancer with increasing dietary cadmium intake, adjusted for total energy intake, smoking history, consumption of vegetables, potatoes, and whole grains, multivitamin use, education, race, body mass index, physical activity, age at first birth, postmenopausal hormone use, and mammography. RESULTS: Vegetables and grains together contributed an average of 66 % of estimated dietary cadmium. During a mean of 7.5 years of follow-up, 1,026 invasive postmenopausal breast cancers were identified. Among 899 cases with complete covariate information, no evidence of an association between dietary cadmium intake and breast cancer risk was observed (aHR (95 % CI), highest to lowest quartile cadmium: 1.00 (0.72-1.41), p (trend) = 0.95). No evidence was found for interactions between dietary cadmium and breast cancer risk factors, smoking habits, or total intake of calcium, iron, or zinc from diet, supplements, and multivitamins. CONCLUSIONS: This study does not support the hypothesis that dietary cadmium intake is a risk factor for breast cancer. However, non-differential measurement error in the estimate of cadmium intake is likely the most important factor that could have obscured an association.