Supplementary Table 1: Associations Between Breast Cancer Risk, MTHFR and MTR Single Nucleotide

Supplementary Table 1: Associations between breast cancer risk, MTHFR and MTR single nucleotide polymorphisms (SNPs), stratified by quartiles of the principal component analysis (PCA)-derived Mediterranean dietary patterna in the MASTOS study

Supplementary Table 1 (continued): Associations between breast cancer risk, MTHFR and MTR single nucleotide polymorphisms (SNPs), stratified by quartiles of the principal component analysis (PCA)-derived Mediterranean dietary patterna in the MASTOS study

The P-values (<0.05), which are statistically significant, are presented in bold. The odds ratios (ORs) (95% confidence interval (CI)) which correspond to the statistically significant P-values are also presented in bold

aPCA-derived dietary pattern with high loadings of vegetables, fruit, legumes and fish, thus closely resembling the Mediterranean diet

bQuartiles of adherence to PCA-derived Mediterranean dietary pattern, with subjects in quartile 1, showing the lowest consumption of vegetables, fruit, legumes and fish and thus, lowest adherence to this dietary pattern and with subjects in quartile 4, showing the highest consumption of the same four food groups and therefore, highest adherence to this dietary pattern

cAdjusted for menopausal status, age and other PCA-derived dietary components (Patterns 1, 2 and 3)

dP for trend within each quartile of adherence to the PCA-derived Mediterranean dietary pattern across the three genotypes of each SNP

Supplementary Table 2: Associations between breast cancer risk and principal component analysis (PCA)-derived Mediterranean dietary patterna, stratified by genotypes of MTHFR and MTR single nucleotide polymorphisms (SNPs) in the MASTOS study

The P-values (<0.05), which are statistically significant, are presented in bold. The odds ratios (ORs) (95% confidence interval (CI)) which correspond to the statistically significant P-values are also presented in bold

aPCA -derived dietary pattern with high loadings of vegetables, fruit, legumes and fish (continuous values), thus closely resembling the Mediterranean diet

bAdjusted for menopausal status, age and other PCA-derived dietary components (Patterns 1, 2 and 3). The data of the PCA-derived dietary patterns 1, 2 and 3 was treated as continuous variables

Supplementary Table 3: Quantities of one-carbon metabolism micronutrients in a serving of 100g of the foods loaded in the principal component analysis (PCA)-derived dietary patterns 1a, 2b, 3c and 4d in the MASTOS study

Supplementary Table 3 (continued): Quantities of one-carbon metabolism micronutrients in a serving of 100g of the foods loaded in the principal component analysis (PCA)-derived dietary patterns 1a, 2b, 3c and 4d in the MASTOS study

Supplementary Table 3 (continued): Quantities of one-carbon metabolism micronutrients in a serving of 100g of the foods loaded in the principal component analysis (PCA)-derived dietary patterns 1a, 2b, 3c and 4d in the MASTOS study

Supplementary Table 3 (continued): Quantities of one-carbon metabolism micronutrients in a serving of 100g of the foods loaded in the principal component analysis (PCA)-derived dietary patterns 1a, 2b, 3c and 4d in the MASTOS study

aPCA-derived dietary pattern with high loadings of meat and potatoes (Demetriou et al. 2012)

bPCA-derived dietary pattern with high loadings of cereals, milk and dairy products (Demetriou et al. 2012)

cPCA-derived dietary pattern with high loadings of cakes, sweets, nuts, crackers, pasta, and rice (Demetriou et al. 2012)

dPCA-derived dietary pattern with high loadings of fish, fruit, legumes and vegetables, thus closely resembling the Mediterranean diet (Demetriou et al. 2012)

eQuantities for each of the one-carbon metabolism micronutrient of interest (dietary folate, Vitamin B2, B6, B12 and choline) were obtained from the online USDA FNDDS site "What's in the Foods You Eat? Search Tool", which calculates quantities of nutrients in micrograms (mcg) or milligrams (mg) for several measures (e.g grams) of various servings for the foods in the database: http://reedir.arsnet.usda.gov/codesearchwebapp/(qf1nsd5505in0m553j1oqvfb)/codesearch.aspx

More general foods [e.g. vegetables (not specified as to type, cooked, not specified as to fat added in cooking)] and foods which correspond more to the dietary habits of the Cypriot population were chosen from the database in order to better represent each food group loaded in the PCA-derived dietary patterns, since the 32-item food frequency questionnaire (FFQ) of the study did not include detailed food items.