Fig. 1

Volcano plot showing results from the nutrient-wide association study method to evaluate the association between dietary intake of 92 foods and nutrients and breast cancer risk in the EPIC study. The y-axis shows the negative log₁₀ transformation of the estimated q values from the multivariable adjusted Cox proportional hazards regression coefficients, and the x-axis is the estimated log hazard ratio for a one standard deviation increment in intake in relation to risk of breast cancer. The q values represent the adjusted p values using the false discovery rate method, and the horizontal line indicates the false discovery rate threshold of 0.05. Each dietary factor was analysed one at a time, and ordered left to right according to the lowest to highest HR. Models were stratified by age at recruitment and study centre and adjusted for energy intake, history of diabetes, educational attainment, smoking status, BMI, physical activity, menopausal status at baseline, menopausal status by BMI interaction, age at menarche, and the interaction of parous (yes/no) and age at first pregnancy. The six dietary factors that were selected for confirmation in the NLCS are labelled