作者: Julie R. Palmer , Hanna Gerlovin , Traci N. Bethea , Kimberly A. Bertrand , Michael F. Holick
DOI: 10.1186/S13058-016-0745-X
关键词:
摘要: Vitamin D deficiency, which has been linked to an increased risk of colorectal cancer, is particularly common among African Americans. Previous studies vitamin status and breast cancer risk, mostly conducted in white women, have had conflicting results. We examined the relationship between predicted incidence a cohort 59,000 American women. Participants Black Women’s Health Study followed by biennial mail questionnaires since 1995, with self-reported diagnoses confirmed hospital registry records. Repeated five-fold cross-validation linear regression was used derive best 25-hydroxyvitamin (25(OH)D) prediction model based on measured 25(OH)D plasma specimens obtained from 2856 participants 2013–2015 questionnaire-based variables same time frame. In full cohort, including 1454 cases incident invasive Cox proportional hazards models were compute rate ratio (IRR) for each quartile score relative highest quartile. Predicted two-year exposure period cumulative average scores all periods up that time. Twenty-two percent women categorized as “deficient” (<20 ng/mL) another 25 % “insufficient” (20–29 ng/mL). The explained variation correlation coefficient versus observed averaged across runs 0.49 (SD 0.026). Breast decreasing 25(OH)D, p trend 0.015; IRR lowest 1.23 (95 % confidence interval 1.04, 1.46). prospective data, estimated 23 % those relatively high levels. Preventing deficiency may be effective means reducing