This is the peer reviewed version of the following article: Low adherence to the western and high adherence to the mediterranean dietary patterns could prevent colorectal cancer Adela Castelló, Pilar Amiano. Nerea Fernández de Larrea, Vicente Martín, Maria Henar Alonso, Gemma Castaño-Vinyals, Beatriz Pérez-Gómez, Rocío Olmedo- Requena, Marcela Guevara, Guillermo Fernandez-Tardon, Trinidad Dierssen-Sotos, Cristobal Llorens-Ivorra, Jose María Huerta, Rocío Capelo, Tania Fernández-Villa, Anna Díez-Villanueva, Carmen Urtiaga, Jesús Castilla, Jose Juan Jiménez-Moleón, Víctor Moreno, Verónica Dávila-Batista, Manolis Kogevinas, Nuria Aragonés, Marina Pollán, On behalf of MCC-Spain researchers Eur J Nutr. 2019 Jun;58(4):1495-1505. which has been published in final form at https://doi.org/10.1007/s00394-018-1674-5 1 AUTHORS LIST: Adela Castelló1,2,3, Pilar Amiano2,4,5, Nerea Fernández de Larrea1,2, Vicente 1 Martín6, Maria Henar Alonso7,8, Gemma Castaño-Vinyals2,9,10,11, Beatriz Pérez-Gómez1,2,, 2 Rocío Olmedo-Requena2,12,13, Marcela Guevara2,14, Guillermo Fernandez-Tardon2,15, Trinidad 3 Dierssen-Sotos16, Cristobal Llorens-Ivorra17, Jose M. Huerta2,18, Rocío Capelo19, Tania 4 Fernández-Villa6, Anna Díez-Villanueva7, Carmen Urtiaga5, Jesús Castilla2,14, Jose J. Jiménez-5 Moleón2,12,13, Víctor Moreno7,8, Verónica Dávila-Batista6, Manolis Kogevinas2,9,10,11, Nuria 6 Aragonés1,2, Marina Pollán1,2 on behalf of MCC-Spain researchers 7 TITLE: Low adherence to the Western and high adherence to the Mediterranean dietary 8 patterns could prevent colorectal cancer. 9 AFFILIATIONS: 10 1 Cancer Epidemiology Unit, National Center for Epidemiology, Carlos III Institute of Health. 11 Avenida Monforte de Lemos 5, 28029, Madrid, Spain. 12 2 Consortium for Biomedical Research in Epidemiology & Public Health (CIBERESP), Carlos 13 III Institute of Health. Avenida Monforte de Lemos 5, 28029, Madrid, Spain. 14 3 Faculty of Medicine, University of Alcalá, Campus Universitario - C/ 19, Av. de Madrid, 15 Km 33,600, 28871 Alcalá de Henares, Madrid 16 4 Public Health Department of Gipuzkoa, Government of the Basque Country, Avenida 17 Navarra, 4, 20013, San Sebastián, Spain. 18 5 Biodonostia Research Institute. Paseo Dr Beguiristain s/n, 20014, San Sebastián, Spain. 19 6 The Research Group in Gene-Environment and Health Interactions, Vegazana Campus, 20 University of León. Campus Vegazana, s/n, 24071, León, Spain 21 2 7 Unit of Biomarkers and Susceptibility, Cancer Prevention and Control Program, Catalan 22 Institute of Oncology (ICO) and IDIBELL. Gran Via km 2.7, 08907, L’Hospitalet de Llobregat, 23 Spain. 24 8 Department of Clinical Sciences, Faculty of Medicine, University of Barcelona, Campus de 25 Bellvitge, Pavelló de Govern, Feixa Llarga s/n 08907, L'Hospitalet de Llobregat, Spain. 26 9 ISGlobal, Centre for Research in Environmental Epidemiology (CREAL). Carrer del Doctor 27 Aiguader 88, 08003, Barcelona, Spain. 28 10 Universitat Pompeu Fabra (UPF). Carrer del Doctor Aiguader 88, 08003, Barcelona, Spain 29 11 IMIM (Hospital del Mar Medical Research Institute). Carrer del Doctor Aiguader 88, 30 08003, Barcelona, Spain 31 12 Department of Preventive Medicine and Public Health, University of Granada. Av, de la 32 Investigación, 11, 18016, Granada, Spain. 33 13 Instituto de Investigación Biosanitaria ibs.GRANADA, Complejo Hospitales Universitarios 34 de Granada/Universidad de Granada. Edificio Licinio de la Fuente, Calle Dr. Azpitarte, 4, 35 18012, Granada, Spain. 36 14 Public Health Institute of Navarra. Calle Leyre 15, 31003, Pamplona, Spain 37 15 IUOPA, University of Oviedo. Facultad de Medicina, Planta 7, Campus de El Cristo B, 38 33006, Oviedo, Spain. 39 16 Universidad de Cantabria – IDIVAL. Avenida Cardenal Herrera Oria s/n, 39011, 40 Santander, Spain. 41 3 17 Centro de Salud Pública de Dénia. Consellería de Sanidad Universal y Salud Pública. Plaza 42 Jaime I, 5, 03700, Denia, Spain. 43 18 Department of Epidemiology, Murcia Regional Health Council, IMIB-Arrixaca. C/ Luis 44 Fontes Pagán nº 9 - 1ª planta. C.P.- 30003, Murcia, Spain. 45 19 Centro de Investigación en Salud y Medio Ambiente (CYSMA). Universidad de Huelva. 46 Campus Universitario de El Carmen, 21071, Huelva, Spain. 47 CORRESPONDING AUTHOR: 48 Dr. Adela Castelló 49 Cancer Epidemiology Unit, National Center for Epidemiology. 50 Instituto de Salud Carlos III. 51 Avenida Monforte de Lemos, 5, 28029, Madrid, Spain. 52 ORCID: 0000-0002-1308-9927 53 Phone: +34 91 822 2667. 54 Fax: +34 91 387 7815. 55 e-mail: acastello@isciii.es 56 57 4 ABSTRACT 58 Purpose: To assess if the associations found between three previously identified dietary 59 patterns with breast, prostate and gastric cancer are also observed for colorectal cancer (CRC). 60 Methods: MCC-Spain is a multicase-control study that collected information of 1629 incident 61 cases of CRC and 3509 population-based controls from 11 Spanish provinces. Western, Prudent 62 and Mediterranean data-driven dietary patterns - derived in another Spanish case-control study- 63 were reconstructed in MCC-Spain. Their association with CRC was assessed using mixed 64 multivariable logistic regression models considering a possible interaction with sex. Risk by 65 tumor site (proximal colon, distal colon, and rectum) was evaluated using multinomial 66 regression models. 67 Results: While no effect of the Prudent pattern on CRC risk was observed, a high adherence to 68 the Western dietary pattern was associated with increased CRC risk for both males (ORfourth 69 (Q4)vs.first (Q1)quartile (95%CI):1.45(1.11;1.91)) and females (ORQ4vsQ1(95%CI):1.50 (1.07;2.09)) 70 but seem to be confined to distal colon (ORfourth (Q4)vs.first (Q1)quartile (95%CI):2.02(1.44;2.84)) and 71 rectal (ORQ4vsQ1 (95%CI):1.46(1.05;2.01)) tumors. The protective effect of the Mediterranean 72 dietary pattern against CRC was observed for both sexes (Males: 73 ORQ4vsQ1(95%CI):0.71(0.55;0.92) ; Females: ORQ4vsQ1(95%CI):0.56(0.40;0.77)) and for all 74 cancer sites: proximal colon (ORQ4vsQ1(95%CI):0.70(0.51;0.97)), distal colon 75 (ORQ4vsQ1(95%CI):0.65(0.48;0.89), and rectum (ORQ4vsQ1(95%CI):0.60 (0.45;0.81)). 76 Conclusion: Our results are consistent with most of the associations previously found between 77 these patterns and breast, prostate and gastric cancer risk and indicate that consuming whole 78 fruits, vegetables, legumes, olive oil, nuts and fish and avoiding red and processed meat, refined 79 grains, sweets, caloric drinks, juices, convenience food and sauces might reduce CRC risk. 80 5 KEYWORDS: “Colonic Neoplasms”; “Rectal Neoplasms”; “prevention and control”; 81 “Principal Component Analysis”; “Dietary Patterns”; "Diet"; "Diet, Western"; "Diet, 82 Mediterranean". 83 6 ACKNOWLEDGEMENTS: This work was supported by Carlos III Institute of Health 84 grants (PI12/00488, PI12/00265, PI12/00715, PI12/01270, PI11/01403, PI11/01889, 85 PI11/00226, PI11/01810, PI11/02213, PI09/00773, PI09/01286, PI09/01903, PI09/02078, 86 PI09/01662, PI08/1770, PI08/0533, PI08/1359), Spanish Ministry of Economy and 87 Competitiveness (IJCI-2014-20900), Consejería de Salud de la Junta de Andalucía (PI-0306-88 2011; PI-0571-2009); Catalan Government DURSIgrant (2014SGR647);Instituto de Salud 89 Carlos III, co-funded by FEDER funds –a way to build Europe– PI14-00613; Fundación 90 Marqués de Valdecilla (API 10/09); Acción Transversal del Cancer, approved by the Spanish 91 Ministry Council on October 11, 2007; Red Temática de Investigación del Cáncer del ISCIII 92 (RD12/0036/0036); Junta de Castilla y León (LE22A10-2); Consejería de Salud de la Junta 93 de Andalucía (2009-S0143); Conselleria de Sanitat de la Generalitat Valenciana 94 (AP_061/10); Recercaixa (2010ACUP 00310); Regional Government of the Basque Country; 95 Consejería de Sanidad de la Región de Murcia; European Commission grants (FOOD-CT-96 2006-036224-HIWATE); Spanish Association Against Cancer Scientific Foundation; 97 Fundación Caja de Ahorros de Asturias; University of Oviedo. 98 None of the sponsors intervened in any of the stages of the research. 99 INTRODUCTION 100 The incidence of colorectal cancer (CRC) has increased in Europe in the last decades 101 [1], being the second most diagnosed cancer in 2012 [2]. According to the scientific evidence, 102 40-50% of CRC cases are attributable to modifiable risk factors such as diet, physical activity 103 and body weight [3,4], providing major opportunities for prevention. The current evidence, 104 points to a possible protective effect of foods containing dietary fiber and calcium against CRC 105 [5,6] and a detrimental effect of red and processed meat [6,7] and alcohol consumption [6,8,9]. 106 7 Despite the importance of these findings for individual foods, some authors suggest that 107 the evaluation of the effects of full dietary patterns might be more appropriate, since it allows 108 the exploration of the effect of food and nutrient interactions in disease [10-12]. Many indexes 109 have been developed in the last years that evaluate dietary quality against predefined criteria 110 [13,14] and a recent metaanalysis found an inverse association between a high score for these 111 indexes and cancer mortality and/or incidence [15]. However, these indexes are based on results 112 in the area of cardiovascular disease and they refer to a theoretical diet that do not necessarily 113 reflect the eating habits of a particular population. Moreover, moderate alcohol intake is 114 positively considered in most of these indexes although alcohol consumption as low as one 115 drink per day increases the risk of several tumors, including colorectal cancer [6]. In fact, some 116 authors suggested that the lack of concordance of the results found for diet quality indexes and 117 cancer might be due to their positive scoring for alcohol consumption [16]. As an alternative 118 approach, dietary patterns that accurately represent the diet in a specific population can be 119 identified with statistical methods like principal component analysis. These patterns also 120 present the advantage of being extracted independently of disease associations, which allows 121 exploration of the role of actual dietary habits in different health conditions. The scarce existing 122 results for data-driven dietary patterns and CRC, indicate a possible protective effect of the so 123 called Mediterranean/Healthy/Prudent dietary pattern [17-22] on CRC and a harmful effect of 124 a pattern labelled as Western/Unhealthy diet [17-23], but the evidence is still insufficient [6]. 125 A recent Spanish study on female breast cancer (BC) –EpiGEICAM- identified three 126 data-driven dietary patterns [24] labelled as Western (associated with increased risk of BC), 127 Prudent (not associated with BC) and Mediterranean (protective against BC). EpiGEICAM 128 presents the novelty of being able to identify, over the same population, two different patterns 129 (Prudent and Mediterranean) commonly interchanged in the literature [9,18,20,21,23,25]. 130 However, these patterns do not always represent the same dietary habits and the differences 131 8 might be determinant in their association with disease risk, as it was the case for BC in the 132 EpiGEICAM context [25]. Therefore, the replication of these patterns in different populations 133 and the exploration of their association with tumors other than BC are of great scientific interest. 134 In fact, the reproducibility of the patterns found has already been assessed in a different sample 135 of Spanish women [26] and similar associations have been observed for other tumours and 136 individuals. The detrimental effect of a high adherence to the Western dietary pattern has been 137 corroborated for breast cancer [27] and also observed for gastric [28] cancer. These studies also 138 show different results for the Prudent (null effect) and Mediterranean (protective) patterns in 139 the case of breast [27], prostate [29] and gastric cancer [28]. 140 The objective of the present study is to assess if the associations found between these 141 dietary patterns -Western, Prudent and Mediterranean- with breast [24,27], prostate [29] and 142 gastric cancer [28] risk in our country are also observed for CRC risk, and to evaluate possible 143 differences by sex and cancer site. 144 PATIENTS AND METHODS 145 The multicase-control MCC-Spain study [30] recruited between 2008 and 2013 146 histologically confirmed incident cases of five tumors: breast, prostate, colorectal, stomach and 147 chronic lymphocytic leukemia. Cases were recruited in 23 hospitals from 12 provinces and a 148 single set of population controls, frequency matched by age and sex with the overall distribution 149 of cases in each province, were randomly selected from the lists of residents assigned to primary 150 health-care centers belonging to the same catchment area of each collaborating hospital. MCC-151 Spain recruited 2140 histologically confirmed CRC cases and 3950 population-based controls 152 from 11 of the 12 contributing provinces. These numbers, represented the 64% of the CRC 153 cases and the 53% of controls invited to participate (supplementary material, figure S1). 154 Potential participants had to be able to answer the questionnaire, had tolive in the study area for 155 9 at least 6 months before the diagnosis (cases) or at recruitment (controls) and had to be 20-85 156 years old. Cases were identified as soon as possible after the diagnosis, and histologically 157 confirmed incident cases of colon (ICD10 codes C18: malignant neoplasm of colon and 158 D01.0:Carcinoma in situ of colon) or rectum (ICD10 codes C19: Malignant neoplasm of 159 rectosigmoid junction ; C20: Malignant neoplasm of rectum; D01.1: Carcinoma in situ of 160 rectosigmoid junction and D01.2: Carcinoma in situ of rectum) cancer with no prior history of 161 the disease and diagnosed within the recruitment period were included. They were classified 162 according to the localization of tumor in proximal colon (including caecum, ascending & 163 transverse colon and hepatic and splenic flexures), distal colon (including descending and 164 sigmoid colon) and rectal cancer. When more than one tumor in different locations were 165 diagnosed at the same time, the site in which the tumor was more invasive was assigned. 166 Information on socio-demographic factors, lifestyle and personal/family medical history 167 was collected with a questionnaire administered by trained personnel in a face-to-face 168 interview. Subsequent telephone contact was made to complete missing values on key 169 variables. Height and weight at different ages were self-reported and diet was assessed with a 170 154-items semi-quantitative food frequency questionnaire (FFQ), which was based on an 171 instrument validated in Spain [31]. Dietary information referred to the previous year before 172 diagnosis (cases) or before interview (controls). 173 In the present study, three dietary patterns identified in a previous Spanish case-control 174 study (EpiGEICAM) on female breast cancer (BC) [24] are examined: A Western dietary 175 pattern positively associated with BC risk that is characterized by high intakes of high-fat dairy 176 products, processed meat, refined grains, sweets, caloric drinks, convenience food and sauces 177 and by low intakes of low-fat dairy products and whole grains; A Prudent pattern not related to 178 BC that represented high intakes of low-fat dairy products, vegetables, fruits, whole grains and 179 juices; and a Mediterranean pattern that seemed to be protective and denoted a high intake of 180 10 fish, vegetables, legumes, boiled potatoes, fruits, olives and vegetable oil – mainly olive oil 181 (72%), and olives (22%) in our context-, and a low intake of juices. These three dietary patterns 182 were identified in the EpiGEICAM sample in 2014 by grouping the dietary intake information 183 collected with a 117 FFQ into 26 inter-correlated food groups and applying principal 184 components analysis (PCA) without rotation of the variance-covariance matrix over these 26 185 food groups [32]. This method defines a set of weights (pattern loadings) associated with each 186 food group that represents the correlation between food consumption and the 187 component/pattern scores. Pattern loadings can be used to reproduce such patterns in other 188 samples as explained in detail elsewhere [25,26]. Briefly, we grouped 146 of the 154 items of 189 the MCC-Spain FFQ (excluding non-caloric and alcoholic beverages) into 26 food groups 190 defined in the EpiGEICAM study (see Table 1 for detailed information on the composition of 191 food groups and their weight in the patterns). Afterwards, the scores of adherence to the 192 Western, Prudent and Mediterranean dietary patterns of the MCC-Spain participants were 193 calculated as a linear combination of the pattern loadings for each food group and dietary 194 pattern extracted from the EpiGEICAM study [24] (Table 1) and the food group consumption 195 reported by the MCC-Spain participants. 196 After describing the data, crude and adjusted associations between adherence to each 197 dietary pattern and CRC risk were evaluated using logistic regression models with random 198 province-specific intercepts. As fixed-effects terms, caloric and alcohol intake, self-reported 199 body mass index (BMI) and physical activity (metabolic equivalents (METs)) during the 10 200 years before diagnosis (cases) / interview (controls), age, smoking status, education, family 201 history of CRC and sex were considered as potential confounders. Scores of adherence were 202 analyzed both, as categorical (grouping the scores of adherence into quartiles of their 203 distribution among controls) and continuous variables (1-standard deviation increase taking 204 into account the dispersion among controls). Heterogeneity of the effects by sex was tested 205 11 including in the models an interaction term between the score of adherence and sex. 206 Multinomial logistic regression models were used to evaluate the association of the adherence 207 to the Western, Prudent and Mediterranean dietary patterns with proximal colon, distal colon 208 and rectal cancer separately. These models were adjusted by the same set of variables described 209 before but including province of residence as a fixed effect term. 210 Finally, assuming a causal relationship between the adherence to each of the patterns 211 and CRC risk for all analyses, the population attributable fraction (PAF%) was calculated using 212 Hanley’s J.A. formula [33] to estimate the proportion of total cancer in the population that 213 hypothetically would not have occurred if all participants were in the optimal quartile of 214 adherence to the dietary patterns (first quartile for Western and fourth quartile for Prudent and 215 Mediterranean dietary patterns). Confidence intervals for PAF were computed using bootstrap 216 with 500 iterations. 217 Analyses were performed using STATA/MP (version 14.1, 2015, StataCorp LP) and 218 statistical significance was set at 2-sided p <0.05. 219 RESULTS 220 Among the initially recruited participants, 3509 (89%) controls and 1889 (88%) cases 221 reported data on diet. Cases that provided dietary information later than 6 months after 222 diagnosis were excluded (n=260). Therefore, 1629 cases and 3509 controls aged 22 to 85 years 223 were included in the study (supplementary figure S1). The multivariable analyses were carried 224 out over 1530 cases and 3240 controls because data on either BMI (<5%), physical activity 225 (<1%), smoking status (<1%), total energy (<2%) or alcohol intake (<2%) was missing for 99 226 cases and 269 controls. 227 Compared to controls, CRC cases were more adherent to the Western (p<0.001) and 228 Mediterranean (p=0.015) dietary patterns and reported higher energy (p<0.001) and alcohol 229 12 (p=0.001) intake. CRC cases were also older (p<0.001) and reported higher BMI (p<0.001) and 230 lower levels of physical activity (p<0.001). The proportion of former smokers (p<0.001), males 231 (p<0.001), participants with no formal education (p<0.001) or with family history of CRC 232 (p<0.001) was also higher among cases (Table 2). 233 Results from Table 3 revealed a positive association between a high adherence to the 234 Western dietary pattern and global CRC (ORfourth(Q4)vs.first(Q1)quartile (95%CI):1.50(1.20;1.87)) 235 risk that was similar among males and females (p-interaction=0.615). Once the difference in 236 calorie intake was taken into account, a high adherence to the Mediterranean pattern appeared 237 to be protective (ORQ4vsQ1 (95%CI):0.65(0.53;0.80)), with this effect slightly stronger among 238 females (ORQ4vsQ1 (95%CI):0.56(0.40;0.77)) than among males (ORQ4vsQ1 239 (95%CI):0.71(0.55;0.92)), though the p-value for the heterogeneity of the linear effects was not 240 significant (p-interaction=0.733). Assuming a causal relationship between diet and CRC risk, 241 the estimations indicate that 1/4 and 1/5 of CRC cases could have been prevented if all the 242 participants had been in the lowest category of adherence to the Western and in the highest 243 category of adherence to the Mediterranean dietary patterns respectively. 244 Stratified results by tumor subtype (Table 4) also indicate a detrimental effect of a high 245 adherence to the Western dietary pattern over CRC risk that seems to be confined to distal colon 246 (ORQ4vsQ1 (95%CI):2.02(1.44;2.84)) and rectal tumors (ORQ4vsQ1 (95%CI):1.46(1.05;2.01); p-247 heterogeneity=0.087), while the protective effect of the Mediterranean dietary pattern was 248 similar for all tumor sites (Proximal colon: ORQ4vsQ1 (95%CI):0.70(0.51;0.97); Distal Colon: 249 ORQ4vsQ1 (95%CI):0.65(0.48;0.89); Rectum: ORQ4vsQ1 (95%CI):0.60 (0.45;0.81); p-250 heterogeneity=0.746). In agreement with these results, it was estimated that more than 1/3 of 251 distal colon and 1/4 of rectum tumors could have been prevented if all the study participants 252 were in the lowest quartile of adherence to the Western dietary pattern and 1/5 for distal colon 253 13 and 1/4 for rectum tumors could have been prevented with the highest adherences to 254 Mediterranean dietary pattern. 255 A high adherence to the prudent pattern did not show an association with CRC risk. 256 DISCUSSION 257 The detrimental effect of a high adherence to the Western dietary pattern for breast 258 [24,27] and gastric [28] cancer and the differential effect of a high adherence to the Prudent 259 (null) and to the Mediterranean (protective) dietary patterns over breast [24,27], prostate [29] 260 and gastric cancer [28] identified in the previous studies was also found for CRC in the present 261 work. Concretely, we found that a high adherence to the Western dietary pattern might increase 262 CRC risk in both males and females and that such risk might be confined to distal colon and 263 rectal cancer. Also, a high adherence to the Mediterranean dietary pattern showed a general 264 protective effect against CRC that was very similar among males and females and for all cancer 265 sites. On the contrary, the adherence to the Prudent dietary pattern was not associated to CRC. 266 Some recent reviews and metaanalysis [9,19,22], also report a positive association 267 between a high adherence to the Western dietary pattern and CRC risk and a protective effect 268 of a diet rich in fruits, vegetables, legumes and/or fish. The studies published after these 269 reviews, also report a positive association of a high adherence to the Western dietary pattern 270 with global CRC risk [18,20,21,23] and a possible protective effect of a Healthy diet against 271 this tumor [18,20,21]. In agreement with our results, some authors conclude that the effect of 272 the Western and Healthy diet might be stronger for distal colon and rectal cancer [21,22] or 273 indicate stronger effects of the Western diet in distal colon tumors [9]. Only a few of these 274 studies provide information of a possible interaction between diet and sex [18,20,21] and none 275 of them report significant differences.. Similarly, the current evidence for index based dietary 276 patterns point to a detrimental effect of pro-inflammatory diets (similar to our Western pattern) 277 14 for CRC risk [34] and a protective effect of diets that share common characteristics with our 278 Mediterranean pattern against this type of tumor [34,35]. One of the most important findings of 279 the present study is the difference in the associations found for Prudent and Mediterranean 280 dietary patterns. To understand these differences, we explored the associations of CRC risk 281 with individual food groups (supplementary Table S1). We believe that the protective effect of 282 the Mediterranean pattern against the null effect of the Prudent might be greatly explained by 283 the protective effect of oily fish, nuts and olives and olive oil, only present in the Mediterranean 284 pattern, but also by the detrimental effect of juices intake, only included in the Prudent pattern, 285 that might counteract the positive effect of a high consumption of fruits, vegetables and whole 286 grains characteristic of this pattern. 287 Some biological mechanisms support the associations found. On the one hand, the 288 “Western”-like diet high in fat, refined grains, red and processed meats and sweets has been 289 associated with higher levels of inflammatory markers [36] and with inflammation-related 290 chronic diseases [37]. Moreover, the high content of iron in meat products present in this pattern 291 generates free radicals that attack DNA and damage the tissue [38]. Additionally, processing 292 meat at high temperatures produces carcinogens such as N-nitroso and polycyclic aromatic 293 hydrocarbons [39]. On the other hand, the antioxidants from fruits, vegetables and legumes 294 present in the Mediterranean pattern may reduce risk by quenching free radicals and reducing 295 oxidative damage to DNA [40]. Furthermore, fiber dilutes faecal content, decreases transit time 296 and increases stool weight [41] contributing to a healthier gastrointestinal tract. Different 297 carcinogenic pathways in proximal and distal tumors have been suggested, based on their 298 molecular differences [42]. In this sense, the higher effect of the Western dietary pattern 299 (characterized by a low dietary fiber intake) in distal colon and rectal tumors, might reflect a 300 higher susceptibility to dietary carcinogens due to a less mature phenotype and lower immune 301 activity of dendritic cells involved in immunologic surveillance at this location [43]. Olive oil 302 15 intake has also been suggested to inhibit colon cancer development by inducing apoptosis and 303 down-regulating the expression of cyclooxygenase2 and Bcl-2 proteins that have a crucial role 304 in colorectal carcinogenesis [44]. Finally, the gut microbiome seems to play an important role 305 in colorectal carcinogenesis [45], and dietary habits strongly influence it [46]. Turnbaugh et al. 306 [46] recently demonstrated in an animal model that changing from low-fat, plant based diets to 307 high-fat, high-sugar diets can shift the structure of the microbiota, modify the representation of 308 metabolic pathways in the microbiome, and alter microbiome gene expression . 309 Our results should be interpreted in the context of the study´s limitations. Recall bias is 310 always a concern in case-control studies. Anticipating the existence of this bias, some questions 311 about general dietary habits were included in the questionnaire and used to adjust the responses 312 to the FFQ [47]. Additionally, only cases that responded to the questionnaire within the 6 313 months following their diagnosis were included. On the other hand, the participation rates (64% 314 among CRC cases and 53% among controls) might give rise to some concerns about a possible 315 selection bias. In this sense, participating controls might have healthier lifestyles than the 316 general population, resulting in an overestimation of the effects. However, no effect was found 317 for the prudent pattern that includes consumption of products widely known as “Healthy”. 318 Therefore, we believe that this bias, if it exists, would be non-differential. Finally, the biological 319 plausibility of the associations found, their strength, their consistency across sex and tumor site, 320 their consistency with the results from other studies on CRC [9,17-23] and the reproducibility 321 of the results across different studies and tumors [24,27-29], deem it unlikely that our findings 322 are a result of recall or selection bias. 323 One of the main strengths of the current research is the recruitment of histologically 324 confirmed incident cases of CRC and population controls. Furthermore, the geographical 325 variability of the recruited participants, coming from 11 provinces from the North, South, 326 Center, West and East of the country, ensured the representation of the different diets coexisting 327 16 within Spain. Also, the sample size allowed the evaluation of potential interactions of diet and 328 sex and the exploration of the associations by tumor localization. We also carried out a 329 sensitivity analysis to explore all the associations excluding 42 in situ cases and obtained very 330 similar results that led to the same exact conclusions (supplementary material tables S2 and 331 S3). Additionally, as mentioned before, we explored the associations of CRC risk with 332 individual food groups to ensure the associations found for patterns are not only due to the 333 presence in the pattern of one or two foods associated with this tumor (supplementary material 334 table S1). High consumers of high fat dairy products, meats, refined grains and sweets (products 335 characteristic of the Western Pattern) showed higher risk of CRC, while high consumers of oily 336 fish, vegetables, fruits, nuts and olive oil (foods present in the Mediterranean pattern) seemed 337 to be protected against this tumor. Therefore, most of the components of the two patterns 338 associated with CRC were also individually associated with this tumor, making it unlikely that 339 the associations found for the whole dietary patterns are due only to the association of CRC 340 with some individual foods. Finally, the reproducibility [26] and applicability [25] of the 341 Western, Prudent and Mediterranean dietary patterns applied here was previously tested, 342 demonstrating that the scores of adherence to these patterns can be calculated following the 343 exact same rules over different populations, resulting in different levels of adherence but still 344 being valid, which is supported by the similitude of the results found for breast [24,27], prostate 345 [29] and gastric cancer [28] and the present results found for CRC. 346 Our results provide evidence about very specific associations between diet and CRC 347 that could be useful to clinical practitioners and public health professionals to offer nutritional 348 recommendations based on avoiding the Western dietary pattern and promoting the 349 Mediterranean diet. Even though other risk factors are involved in the genesis of these type of 350 tumors, diet is a key risk factor for colorectal cancer. In this sense, if a country like Spain, with 351 a high compliance with the Mediterranean diet and a moderate adherence to the Western diet, 352 17 can benefit from abandoning the latter in favor of the former, the benefit might be greater in 353 countries with unhealthier diets. 354 355 CONCLUSION 356 A high consumption of fruits, vegetables and whole grains combined with a low dietary 357 fat intake might not be enough to prevent CRC. A fair percentage of colorectal cancer cases 358 could be reduced in the general population by providing dietary recommendations based in a 359 decrease of the consumption of high-fat dairy products, red and processed meat, refined grains, 360 sweets, caloric drinks, juices, convenience food and sauces in favor of an increase in the intake 361 of whole fruits, vegetables, legumes, olive oil, nuts and fish, especially for distal colon and 362 rectal tumors. 363 364 365 366 18 ETHICAL STANDARDS: 367 The MCC-Spain study protocol was approved by the Ethics Committee of each the 368 participating institutions and has been performed in accordance with the ethical standards as 369 laid down in the 1964 Declaration of Helsinki and its later amendments. All participants were 370 informed about the study objectives and signed an informed consent. 371 CONFLICT OF INTEREST: 372 The authors declare that they have no conflict of interest. 373 REFERENCES: 374 1. Arnold M, Karim-Kos HE, Coebergh JW, Byrnes G, Antilla A, Ferlay J, Renehan AG, Forman D, 375 Soerjomataram I (2015) Recent trends in incidence of five common cancers in 26 European countries since 376 1988: Analysis of the European Cancer Observatory. European journal of cancer 51 (9):1164-1187. 377 doi:10.1016/j.ejca.2013.09.002 378 2. Stewart B, Wild C, Editors (2014) (2014) World Cancer Report 2014. Lyon, France: International 379 Agency for Research on Cancer. . 380 3. WCRF (2015) Cancer preventability estimates for diet, nutrition, body fatness, and physical activity. 381 http://www.wcrf.org/int/cancer-facts-figures/preventability-estimates/cancer-preventability-estimates-diet-382 nutrition. Accessed 26/10/2016 2016 383 4. Whiteman DC, Webb PM, Green AC, Neale RE, Fritschi L, Bain CJ, Parkin DM, Wilson LF, Olsen 384 CM, Nagle CM, Pandeya N, Jordan SJ, Antonsson A, Kendall BJ, Hughes MC, Ibiebele TI, Miura K, Peters S, 385 Carey RN (2015) Cancers in Australia in 2010 attributable to modifiable factors: summary and conclusions. 386 Australian and New Zealand journal of public health 39 (5):477-484. doi:10.1111/1753-6405.12471 387 5. Song M, Garrett WS, Chan AT (2015) Nutrients, foods, and colorectal cancer prevention. 388 Gastroenterology 148 (6):1244-1260.e1216. doi:10.1053/j.gastro.2014.12.035 389 6. WCRF/AICR (2017) World Cancer Research Fund / American Institute for Cancer Research. 390 Continuous Update Project Report. Diet, nutrition, physical activity and colorectal cancer. 2017 391 http://www.wcrf.org/sites/default/files/CUP%20Colorectal%20Report_2017_Digital.pdf. Accessed 10th October 392 2017 393 7. Aune D, Chan DS, Vieira AR, Navarro Rosenblatt DA, Vieira R, Greenwood DC, Kampman E, Norat T 394 (2013) Red and processed meat intake and risk of colorectal adenomas: a systematic review and meta-analysis of 395 epidemiological studies. Cancer causes & control : CCC 24 (4):611-627. doi:10.1007/s10552-012-0139-z 396 8. Cai S, Li Y, Ding Y, Chen K, Jin M (2014) Alcohol drinking and the risk of colorectal cancer death: a 397 meta-analysis. European journal of cancer prevention : the official journal of the European Cancer Prevention 398 Organisation (ECP) 23 (6):532-539. doi:10.1097/cej.0000000000000076 399 9. Magalhaes B, Peleteiro B, Lunet N (2012) Dietary patterns and colorectal cancer: systematic review and 400 meta-analysis. European journal of cancer prevention : the official journal of the European Cancer Prevention 401 Organisation (ECP) 21 (1):15-23. doi:10.1097/CEJ.0b013e3283472241 402 10. Barkoukis H (2007) Importance of understanding food consumption patterns. J Am Diet Assoc 107 403 (2):234-236 404 11. Hu FB (2002) Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 405 13 (1):3-9 406 12. Jacques PF, Tucker KL (2001) Are dietary patterns useful for understanding the role of diet in chronic 407 disease? The American journal of clinical nutrition 73 (1):1-2 408 19 13. Fung TT, McCullough ML, Newby PK, Manson JE, Meigs JB, Rifai N, Willett WC, Hu FB (2005) 409 Diet-quality scores and plasma concentrations of markers of inflammation and endothelial dysfunction. The 410 American journal of clinical nutrition 82 (1):163-173 411 14. Liese AD, Krebs-Smith SM, Subar AF, George SM, Harmon BE, Neuhouser ML, Boushey CJ, Schap 412 TE, Reedy J (2015) The Dietary Patterns Methods Project: synthesis of findings across cohorts and relevance to 413 dietary guidance. The Journal of nutrition 145 (3):393-402. doi:10.3945/jn.114.205336 414 15. Schwingshackl L, Bogensberger B, Hoffmann G (2018) Diet Quality as Assessed by the Healthy Eating 415 Index, Alternate Healthy Eating Index, Dietary Approaches to Stop Hypertension Score, and Health Outcomes: 416 An Updated Systematic Review and Meta-Analysis of Cohort Studies. Journal of the Academy of Nutrition and 417 Dietetics 118 (1):74-100.e111. doi:10.1016/j.jand.2017.08.024 418 16. Giacosa A, Barale R, Bavaresco L, Gatenby P, Gerbi V, Janssens J, Johnston B, Kas K, La Vecchia C, 419 Mainguet P, Morazzoni P, Negri E, Pelucchi C, Pezzotti M, Rondanelli M (2013) Cancer prevention in Europe: 420 the Mediterranean diet as a protective choice. European journal of cancer prevention : the official journal of the 421 European Cancer Prevention Organisation (ECP) 22 (1):90-95. doi:10.1097/CEJ.0b013e328354d2d7 422 17. Safari A, Shariff ZM, Kandiah M, Rashidkhani B, Fereidooni F (2013) Dietary patterns and risk of 423 colorectal cancer in Tehran Province: a case-control study. BMC public health 13:222. doi:10.1186/1471-2458-424 13-222 425 18. Park Y, Lee J, Oh JH, Shin A, Kim J (2016) Dietary patterns and colorectal cancer risk in a Korean 426 population: A case-control study. Medicine 95 (25):e3759. doi:10.1097/md.0000000000003759 427 19. Feng YL, Shu L, Zheng PF, Zhang XY, Si CJ, Yu XL, Gao W, Zhang L (2017) Dietary patterns and 428 colorectal cancer risk: a meta-analysis. European journal of cancer prevention : the official journal of the 429 European Cancer Prevention Organisation (ECP) 26 (3):201-211. doi:10.1097/cej.0000000000000245 430 20. Haslam A, Wagner Robb S, Hebert JR, Huang H, Ebell MH (2017) Association between dietary pattern 431 scores and the prevalence of colorectal adenoma considering population subgroups. Nutrition & dietetics: the 432 journal of the Dietitians Association of Australia. doi:10.1111/1747-0080.12400 433 21. Mehta RS, Song M, Nishihara R, Drew DA, Wu K, Qian ZR, Fung TT, Hamada T, Masugi Y, da Silva 434 A, Shi Y, Li W, Gu M, Willett WC, Fuchs CS, Giovannucci EL, Ogino S, Chan AT (2017) Dietary Patterns and 435 Risk of Colorectal Cancer: Analysis by Tumor Location and Molecular Subtypes. Gastroenterology 152 436 (8):1944-1953.e1941. doi:10.1053/j.gastro.2017.02.015 437 22. Randi G, Edefonti V, Ferraroni M, La Vecchia C, Decarli A (2010) Dietary patterns and the risk of 438 colorectal cancer and adenomas. Nutrition reviews 68 (7):389-408. doi:10.1111/j.1753-4887.2010.00299.x 439 23. Tayyem RF, Bawadi HA, Shehadah I, Agraib LM, AbuMweis SS, Al-Jaberi T, Al-Nusairr M, Bani-440 Hani KE, Heath DD (2017) Dietary patterns and colorectal cancer. Clinical nutrition (Edinburgh, Scotland) 36 441 (3):848-852. doi:10.1016/j.clnu.2016.04.029 442 24. Castello A, Pollan M, Buijsse B, Ruiz A, Casas AM, Baena-Canada JM, Lope V, Antolin S, Ramos M, 443 Munoz M, Lluch A, de Juan-Ferre A, Jara C, Jimeno MA, Rosado P, Diaz E, Guillem V, Carrasco E, Perez-444 Gomez B, Vioque J, Boeing H, Martin M (2014) Spanish Mediterranean diet and other dietary patterns and 445 breast cancer risk: case-control EpiGEICAM study. British journal of cancer 111 (7):9. 446 doi:10.1038/bjc.2014.434 447 25. Castello A, Buijsse B, Martin M, Ruiz A, Casas AM, Baena-Canada JM, Pastor-Barriuso R, Antolin S, 448 Ramos M, Munoz M, Lluch A, de Juan-Ferre A, Jara C, Lope V, Jimeno MA, Arriola-Arellano E, Diaz E, 449 Guillem V, Carrasco E, Perez-Gomez B, Vioque J, Pollan M, researchers G (2016) Evaluating the Applicability 450 of Data-Driven Dietary Patterns to Independent Samples with a Focus on Measurement Tools for Pattern 451 Similarity. Journal of the Academy of Nutrition and Dietetics 116 (12):1914-1924 e1916. 452 doi:10.1016/j.jand.2016.05.008 453 26. Castello A, Lope V, Vioque J, Santamarina C, Pedraz-Pingarron C, Abad S, Ederra M, Salas-Trejo D, 454 Vidal C, Sanchez-Contador C, Aragones N, Perez-Gomez B, Pollan M (2016) Reproducibility of data-driven 455 dietary patterns in two groups of adult Spanish women from different studies. The British journal of nutrition 456 116 (4):734-742. doi:10.1017/S000711451600252X 457 27. Castello A, Boldo E, Perez-Gomez B, Lope V, Altzibar JM, Martin V, Castano-Vinyals G, Guevara M, 458 Dierssen-Sotos T, Tardon A, Moreno V, Puig-Vives M, Llorens-Ivorra C, Alguacil J, Gomez-Acebo I, Castilla J, 459 Gracia-Lavedan E, Davila-Batista V, Kogevinas M, Aragones N, Amiano P, Pollan M (2017) Adherence to the 460 Western, Prudent and Mediterranean dietary patterns and breast cancer risk: MCC-Spain study. Maturitas 103:8-461 15. doi:10.1016/j.maturitas.2017.06.020 462 28. Castelló A, Fernández de Larrea N, Martin M, Dávila-Batista V, Boldo E, Guevara M, Moreno V, 463 Castaño-Vinyals G, Gómez-Acebo I, Fernandez-Tardon G, Peiro R, Olmedo-Requena R, Capelo R, Navarro C, 464 Pacho-Valbuena S, Pérez-Gómez B, Kogevinas M, Pollán M, Aragonés N (2017) High Adherence to the 465 Western, Prudent and Mediterranean Dietary Patterns and Risk of Gastric Adenocarcinoma. (In Press). Gastric 466 Cancer 467 20 29. Castello A, Boldo E, Amiano P, Castano-Vinyals G, Aragones N, Gomez-Acebo I, Peiro R, Jimenez-468 Moleon JJ, Alguacil J, Tardon A, Cecchini L, Lope V, Dierssen-Sotos T, Mengual L, Kogevinas M, Pollan M, 469 Perez-Gomez B, researchers M-S (2017) Mediterranean dietary pattern is associated to low risk of aggressive 470 prostate cancer: MCC-Spain study. The Journal of urology. doi:10.1016/j.juro.2017.08.087 471 30. Castano-Vinyals G, Aragones N, Perez-Gomez B, Martin V, Llorca J, Moreno V, Altzibar JM, Ardanaz 472 E, de Sanjose S, Jimenez-Moleon JJ, Tardon A, Alguacil J, Peiro R, Marcos-Gragera R, Navarro C, Pollan M, 473 Kogevinas M, Group MC-SS (2015) Population-based multicase-control study in common tumors in Spain 474 (MCC-Spain): rationale and study design. Gaceta sanitaria / SESPAS 29 (4):308-315. 475 doi:10.1016/j.gaceta.2014.12.003 476 31. Garcia-Closas R, Garcia-Closas M, Kogevinas M, Malats N, Silverman D, Serra C, Tardon A, Carrato 477 A, Castano-Vinyals G, Dosemeci M, Moore L, Rothman N, Sinha R (2007) Food, nutrient and heterocyclic 478 amine intake and the risk of bladder cancer. European journal of cancer 43 (11):1731-1740. 479 doi:10.1016/j.ejca.2007.05.007 480 32. Burt C (1948) Factor Analysis and canonical correlations. Br J Math Stat Psychol 1 (2):95-106 481 33. Hanley JA (2001) A heuristic approach to the formulas for population attributable fraction. Journal of 482 epidemiology and community health 55 (7):508-514 483 34. Steck SE, Guinter M, Zheng J, Thomson CA (2015) Index-based dietary patterns and colorectal cancer 484 risk: a systematic review. Advances in nutrition (Bethesda, Md) 6 (6):763-773. doi:10.3945/an.115.009746 485 35. Schwingshackl L, Schwedhelm C, Galbete C, Hoffmann G (2017) Adherence to Mediterranean Diet 486 and Risk of Cancer: An Updated Systematic Review and Meta-Analysis. Nutrients 9 (10). 487 doi:10.3390/nu9101063 488 36. Barbaresko J, Koch M, Schulze MB, Nothlings U (2013) Dietary pattern analysis and biomarkers of 489 low-grade inflammation: a systematic literature review. Nutrition reviews 71 (8):511-527. 490 doi:10.1111/nure.12035 491 37. Thorburn AN, Macia L, Mackay CR (2014) Diet, metabolites, and "western-lifestyle" inflammatory 492 diseases. Immunity 40 (6):833-842. doi:10.1016/j.immuni.2014.05.014 493 38. Ashmore JH, Rogers CJ, Kelleher SL, Lesko SM, Hartman TJ (2016) Dietary Iron and Colorectal 494 Cancer Risk: A Review of Human Population Studies. Critical reviews in food science and nutrition 56 495 (6):1012-1020. doi:10.1080/10408398.2012.749208 496 39. Santarelli RL, Pierre F, Corpet DE (2008) Processed meat and colorectal cancer: a review of 497 epidemiologic and experimental evidence. Nutrition and cancer 60 (2):131-144. 498 doi:10.1080/01635580701684872 499 40. Fang YZ, Yang S, Wu G (2002) Free radicals, antioxidants, and nutrition. Nutrition (Burbank, Los 500 Angeles County, Calif) 18 (10):872-879 501 41. Giovannucci E, Willett WC (1994) Dietary factors and risk of colon cancer. Annals of medicine 26 502 (6):443-452 503 42. Missiaglia E, Jacobs B, D'Ario G, Di Narzo AF, Soneson C, Budinska E, Popovici V, Vecchione L, 504 Gerster S, Yan P, Roth AD, Klingbiel D, Bosman FT, Delorenzi M, Tejpar S (2014) Distal and proximal colon 505 cancers differ in terms of molecular, pathological, and clinical features. Annals of oncology : official journal of 506 the European Society for Medical Oncology 25 (10):1995-2001. doi:10.1093/annonc/mdu275 507 43. Bernardo D, Durant L, Mann ER, Bassity E, Montalvillo E, Man R, Vora R, Reddi D, Bayiroglu F, 508 Fernandez-Salazar L, English NR, Peake ST, Landy J, Lee GH, Malietzis G, Siaw YH, Murugananthan AU, 509 Hendy P, Sanchez-Recio E, Phillips RK, Garrote JA, Scott P, Parkhill J, Paulsen M, Hart AL, Al-Hassi HO, 510 Arranz E, Walker AW, Carding SR, Knight SC (2016) Chemokine (C-C Motif) Receptor 2 Mediates Dendritic 511 Cell Recruitment to the Human Colon but Is Not Responsible for Differences Observed in Dendritic Cell 512 Subsets, Phenotype, and Function Between the Proximal and Distal Colon. Cellular and molecular 513 gastroenterology and hepatology 2 (1):22-39 e25. doi:10.1016/j.jcmgh.2015.08.006 514 44. Pelucchi C, Bosetti C, Negri E, Lipworth L, La Vecchia C (2011) Olive oil and cancer risk: an update of 515 epidemiological findings through 2010. Current pharmaceutical design 17 (8):805-812 516 45. Schwabe RF, Jobin C (2013) The microbiome and cancer. Nature reviews Cancer 13 (11):800-812. 517 doi:10.1038/nrc3610 518 46. Turnbaugh PJ, Ridaura VK, Faith JJ, Rey FE, Knight R, Gordon JI (2009) The effect of diet on the 519 human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice. Science translational medicine 520 1 (6):6ra14. doi:10.1126/scitranslmed.3000322 521 47. Calvert C, Cade J, Barrett JH, Woodhouse A (1997) Using cross-check questions to address the 522 problem of mis-reporting of specific food groups on Food Frequency Questionnaires. UKWCS Steering Group. 523 United Kingdom Women's Cohort Study Steering Group. European journal of clinical nutrition 51 (10):708-712 524 525 526 21 Table 1: Composition of food groups based on the food frequency questionnaire of the MCC-527 Spain study and component loadings for each pattern identified in the EpiGEICAM study25. 528 FOOD GROUP FOODa Westb Prudb Medb HIGH FAT DAIRY Whole-fat milk, double cream, condensed milk, whole-fat yogurt, semi-cured, cured, or creamy cheese, blue cheesec, custard, milk shakec, ice-cream, 0.60 -0.11 0.20 LOW FAT DAIRY Semi-skimmed and skimmed milk, soy milkc, skimmed yogurt, curd, cottage or fresh white cheese. -0.49 0.60 -0.01 EGGS Eggs. 0.19 0.08 0.16 WHITE MEAT Chicken, rabbit and duck. 0.08 0.17 0.18 RED MEAT Pork, beef, lamb, liver (beef, pork or chicken), entrails, hamburgers (pork or beef) and meatballs (pork or beef)c. 0.27 0.09 0.22 PROCESSED MEAT Sausages, serrano hamc and other cold meat, bacon, pâté, foie-gras. 0.36 0.10 0.26 WHITE FISH Fresh or frozen white fish (hake, sea bass, sea bream), ½·salted fishc and ½·smoked fishc. 0.01 0.24 0.34 OILY FISH Fresh or frozen blue fish (tuna, swordfish, sardines, anchovies, salmon), canned fish, ½·salted fishc and ½·smoked fishc. 0.05 0.24 0.44 SEAFOOD/SHELL FISH Clams, mussels, oysters, squid, cuttlefish, octopus, prawn, crab, shrimp and similar products. 0.17 0.27 0.35 LEAFY VEGETABLES Spinach, chard, lettuce and other leafy vegetables. -0.11 0.34 0.40 FRUITING VEGETABLES Tomato, eggplant, zucchini, cucumber, pepper, artichoke and avocadoc. 0.00 0.36 0.45 ROOT VEGETABLES Carrot, pumpkin and radishc. 0.05 0.35 0.44 OTHER VEGETABLES Cooked cabbage, cauliflower or broccoli, onion, green beans, asparagus, mushroomsc, corn, garlic, gazpachoc, vegetable soupc and other vegetablesc. -0.04 0.40 0.42 LEGUMESd Peasc, lentilsc, chickpeasc, beansc and broad beansc. 0.21 0.15 0.34 POTATOES Roasted or boiled potatoes and sweet potatoesc. 0.17 0.25 0.40 FRUITS Orange, grapefruitc, mandarin, banana, apple, pear, grapes, kiwi, strawberriesc, cherriesc, peach, figsc, melon or watermelon, prunes, mangoc and papayac and other fresh or dried fruitsc. -0.07 0.31 0.31 NUTS Almonds, peanuts, pine nuts, hazelnut 0.18 0.22 0.29 22 FOOD GROUP FOODa Westb Prudb Medb REFINED GRAINS White-flour bread, rice, pasta 0.37 0.15 0.23 WHOLE GRAINS Whole-grain bread and breakfast cereals -0.43 0.47 -0.06 OLIVES AND VEGETABLE OIL Olives, added olive oil to salads, bread and dishes, other vegetable oils (sunflower, corn, and soybean). 0.12 0.19 0.34 OTHER EDIBLE FATS Margarine, butter and lardc. 0.22 0.02 0.11 SWEETS Chocolate and other sweets, cocoa powder, plain cookies, chocolate cookies, pastries (croissant, donut, cake, pie or similar) 0.35 0.18 0.05 SUGARY Jam, honey, sugar and fruit in sugar syrupc. 0.24 0.05 0.00 JUICES Tomato juicec, freshly squeezed orange juice, juice (other than freshly squeezed) 0.25 0.67 -0.39 CALORIC DRINKS Sugar-sweetened soft drinks and nut milkc. 0.74 0.21 -0.25 CONVENIENCE FOOD Croquette, fish sticks, dumplingsc, kebabc, fried potatoes, crisps, pizza, instant soupc, mayonnaise, tomato sauce, hot saucec, ketchup and other saucesc. 0.47 0.12 0.24 AND SAUCES a Log-transformed centered intake in grams. 529 b West: Western; Prud: Prudent; Med: Mediterranean 530 c in bold items that are included in the FFQ from MCC-Spain study that were not included in 531 EpiGEICAM. 532 d FFQ questionnaire from EpiGEICAM only included a single general question on legumes 533 intake whereas MCC-Spain included more detailed information on the type of legumes. 534 535 23 Table 2. Description of scores of adherence to Western, Prudent and Mediterranean dietary 536 patterns and other baseline characteristics for colorectal cancer cases and controls. 537 Controls Cases p n=3509 n=1629 Western mean (sd)a -0.38 (3.52) 0.14 (3.52) <0.001 Prudent mean (sd)a -0.09 (3.29) -0.19 (3.32) 0.353 Mediterranean mean (sd)a -0.02 (2.89) 0.19 (2.75) 0.015 Energy (kcal/day) mean (sd) 1903.81 (570.75) 2008.24 (638.31) <0.001 Alcohol (g/day) median (IQR) 7.22 (0.00;23.21) 9.38 (0.00;34.72) 0.001 BMIb (kg/m2) mean (sd) 26.61 (4.41) 27.59 (4.46) <0.001 Physical activity (METsb/week) n (%c) <0.001 0 1341 (38%) 855 (52%) 0.1-8 506 (14%) 183 (11%) 8-15.9 422 (12%) 135 (8%) >=16 1202 (34%) 456 (28%) Unknown 38 (1%) 0 (0%) Age (years) mean (sd) 63.20 (11.69) 67.09 (10.63) <0.001 Smoking n (%c) <0.001 Never Smoker 1549 (44%) 680 (42%) Former Smoker 1224 (35%) 660 (41%) Current Smoker 724 (21%) 279 (17%) Unknown 12 (0%) 10 (1%) Education n (%c) <0.001 No formal Education 619 (18%) 522 (32%) Primary School 1143 (33%) 648 (40%) Secondary School 1010 (29%) 311 (19%) University or more 737 (21%) 148 (9%) Previous history of CRCb n (%c) <0.001 No 3101 (88%) 1295 (79%) 2nd Degree 107 (3%) 62 (4%) One of 1st degree 281 (8%) 231 (14%) More than one of 1st degree 20 (1%) 41 (3%) Sex <0.001 Male 1813 (52%) 1043 (64%) Female 1696 (48%) 586 (36%) a The pairwise Pearson correlations for the level of adherence to the three identified dietary 538 patterns were 0.329 for the Western and Prudent, 0.231 for the Western and Mediterranean and 539 0.485 for the Prudent and Mediterranean. 540 b BMI: Body mass index; CRC: Colorectal cancer; METS: Metabolic equivalent. 541 c Percentages might not add up 100 because of rounding. 542 24 Table 3. Association between colorectal cancer incidence and the scores of adherence to Western, Prudent and Mediterranean dietary patterns and 543 attributable fractions for all participants and by sex. 544 ALL MALE FEMALE n=4770 n=2688 n=2082 Co/Caa ORb (95%CI) Co/Caa aORc (95%CI) Co/Caa aORc (95%CI) Co/Caa aORc (95%CI) p-int WESTERN Quartiles Q1 877/322 1 772/292 1 335/160 1 437/132 1 Q2 878/409 1.27 (1.06;1.51) 824/390 1.29 (1.06;1.57) 405/227 1.16 (0.89;1.52) 419/163 1.46 (1.10;1.94) Q3 877/423 1.36 (1.14;1.62) 831/401 1.43 (1.17;1.75) 449/260 1.33 (1.02;1.73) 382/141 1.56 (1.16;2.10) Q4 877/475 1.47 (1.23;1.75) 813/447 1.50 (1.20;1.87) 511/341 1.45 (1.11;1.91) 302/106 1.50 (1.07;2.09) p-trend <0.001 <0.001 0.004 0.009 1SD-increase 1.16 (1.09;1.24) 1.19 (1.10;1.30) 1.21 (1.09;1.34) 1.17 (1.04;1.31) 0.615 PAFd% 24% (12%;36%) 21% (5%;37%) 18% (3%;33%) PRUDENT Quartiles Q1 878/440 1 783/398 1 485/292 1 298/106 1 Q2 876/384 0.83 (0.70;0.98) 811/362 0.87 (0.72;1.05) 430/235 0.84 (0.66;1.05) 381/127 0.95 (0.69;1.31) Q3 877/403 0.89 (0.75;1.05) 827/389 1.00 (0.83;1.21) 412/241 0.94 (0.74;1.19) 415/148 1.13 (0.83;1.54) Q4 878/402 0.88 (0.74;1.04) 819/381 0.94 (0.76;1.15) 373/220 0.88 (0.68;1.13) 446/161 1.05 (0.77;1.44) p-trend 0.242 0.875 0.475 0.515 1SDa-increase 0.96 (0.90;1.02) 0.97 (0.90;1.05) 0.95 (0.86;1.04) 1.02 (0.90;1.15) 0.330 PAFd% 2% (-12%;15%) 4% (-12%;21%) 3% (-12%;19%) MEDITERRANEAN Quartiles Q1 878/394 1 796/359 1 398/206 1 398/153 1 Q2 877/412 0.98 (0.83;1.17) 821/386 0.91 (0.75;1.10) 390/236 0.99 (0.77;1.27) 431/150 0.83 (0.63;1.10) Q3 876/371 0.80 (0.67;0.96) 815/357 0.72 (0.59;0.87) 425/219 0.71 (0.55;0.92) 390/138 0.74 (0.55;0.99) Q4 878/452 0.90 (0.76;1.07) 808/428 0.65 (0.53;0.80) 487/327 0.71 (0.55;0.92) 321/101 0.56 (0.40;0.77) p-trend 0.073 0.000 0.001 0.000 1SDa-increase 0.98 (0.92;1.05) 0.87 (0.80;0.94) 0.88 (0.79;0.96) 0.85 (0.76;0.96) 0.733 PAFd% 20% (8%;33%) 15% (2%;29%) 18% (3%;33%) 25 a Co: Controls; Ca: Cases; SD: Standard deviation. 545 b Crude odds ratio of colorectal cancer associated with the adherence to the Western, Prudent and Mediterranean dietary patterns 546 c Odds ratio of colorectal cancer associated with the adherence to the Western, Prudent and Mediterranean dietary patterns adjusted by sex, age, 547 education, BMI, family history of colorectal cancer, physical activity, smoking status, caloric intake and alcohol intake as fixed effects and 548 province of residence as a random effect. 549 c Same as b including an interaction term with sex. 550 d PAF= Population attributable fraction. Proportion of colorectal cancer cases that could be prevented if all participants were in the most 551 beneficial category of adherence to each pattern (Q1 for Western and Q4 for Prudent and Mediterranean) 552                 11 2 2 3 3 4 4 1 2 41 2 3 3 4 1 1 1 1 1 [ 1 1 1 1 ] PF=Proportion of population in the specific exposure c 100 ategory OR= Odds ratio for the especific Q Q Q Q Q Q Q Q Q Q Q Q Q Q QQ PF OR PF OR PF OR PF OR PF OR PF OR P PAF F OR PF OR                          exposure category 553 26 SUPPLEMENTARY MATERIAL Figure S1. Flow of colorectal cancer cases and controls through the MCC-Spain study stages. CRC, Colorectal cancer. 27 Table S1: Odds Ratio of colorectal cancer associated to quartiles of consumption of 26 food groups not adjusting and adjusting by the consumption of the rest of the foods. 2nd Quartile 3rd Quartile 4th Quartile 2nd Quartile 3rd Quartile 4th Quartile ORa(95%CI) ORa(95%CI) ORa(95%CI) p for trenda ORb(95%CI) ORb(95%CI) ORb(95%CI) p for trendb High fat dairyc 1.13(0.93;1.37) 1.21(1.00;1.48) 1.39(1.15;1.70) 0.001 1.08(0.88;1.32) 1.11(0.89;1.37) 1.19(0.94;1.50) 0.169 Low fatdairyc 0.81(0.68;0.98) 0.86(0.72;1.04) 0.70(0.58;0.85) 0.001 0.88(0.72;1.08) 0.99(0.8;1.23) 0.86(0.68;1.08) 0.394 Eggsd . 1.13(0.95;1.34) 1.46(1.18;1.82) 0.003 . 1.00(0.84;1.20) 1.22(0.96;1.53) 0.261 White meatc 1.02(0.83;1.25) 1.12(0.93;1.36) 1.30(1.08;1.57) 0.003 0.98(0.79;1.21) 1.06(0.87;1.30) 1.30(1.06;1.59) 0.005 Red meatc 1.15(0.94;1.40) 1.53(1.26;1.87) 1.65(1.35;2.03) 0.000 1.11(0.90;1.37) 1.42(1.15;1.75) 1.39(1.12;1.74) 0.001 Processed meatc 1.14(0.94;1.40) 1.10(0.90;1.35) 1.58(1.28;1.94) 0.000 1.12(0.91;1.39) 1.02(0.82;1.27) 1.49(1.19;1.86) 0.001 White fishc 1.07(0.88;1.29) 1.10(0.91;1.32) 1.14(0.94;1.38) 0.173 1.07(0.88;1.32) 1.15(0.94;1.41) 1.32(1.07;1.63) 0.009 Oily fishc 0.96(0.80;1.15) 0.81(0.67;0.97) 0.78(0.65;0.95) 0.003 1.01(0.83;1.22) 0.84(0.69;1.03) 0.79(0.64;0.98) 0.014 Seafoodc 1.01(0.84;1.22) 0.90(0.74;1.09) 0.86(0.71;1.05) 0.076 0.98(0.81;1.19) 0.87(0.71;1.06) 0.83(0.67;1.03) 0.049 Leafy vegetablesc 0.90(0.75;1.08) 0.73(0.60;0.88) 0.56(0.46;0.68) 0.000 1.01(0.82;1.23) 0.90(0.73;1.12) 0.75(0.59;0.96) 0.017 Fruiting Vegetablesc 0.77(0.64;0.92) 0.62(0.51;0.75) 0.59(0.48;0.71) 0.000 0.80(0.66;0.98) 0.73(0.59;0.90) 0.86(0.68;1.09) 0.155 Root Vegetablesc 0.86(0.72;1.03) 0.75(0.62;0.90) 0.61(0.50;0.74) 0.000 0.90(0.74;1.08) 0.83(0.68;1.01) 0.76(0.61;0.95) 0.014 Other Vegetablesc 0.92(0.77;1.10) 0.70(0.58;0.84) 0.64(0.53;0.78) 0.000 1.00(0.82;1.23) 0.82(0.66;1.02) 0.87(0.68;1.10) 0.081 Legumesd . 0.87(0.73;1.03) 0.94(0.79;1.11) 0.375 . 0.91(0.76;1.08) 0.99(0.82;1.19) 0.833 Potatoesc 1.27(1.03;1.56) 1.61(1.32;1.97) 1.64(1.34;2.01) 0.000 1.29(1.04;1.60) 1.61(1.30;1.98) 1.65(1.33;2.04) 0.000 Fruitsc 0.93(0.77;1.13) 0.85(0.70;1.03) 0.64(0.52;0.78) 0.000 1.03(0.85;1.26) 0.99(0.8;1.22) 0.80(0.64;1.00) 0.042 Nutsd . 0.94(0.81;1.09) 0.73(0.60;0.89) 0.008 . 1.00(0.85;1.18) 0.90(0.73;1.10) 0.528 Olives and Vegetable Oilc 0.87(0.69;1.08) 0.86(0.73;1.02) 0.79(0.65;0.96) 0.019 0.90(0.72;1.14) 0.92(0.77;1.10) 0.90(0.73;1.12) 0.259 Other Edible Fatsc . 0.96(0.79;1.18) 1.16(0.98;1.36) 0.157 . 0.90(0.72;1.11) 1.02(0.86;1.22) 0.981 Refined Grainsc 1.34(1.09;1.63) 1.37(1.12;1.68) 1.42(1.13;1.78) 0.004 1.20(0.96;1.49) 1.17(0.93;1.47) 1.24(0.96;1.6) 0.245 Whole grainsd . 0.79(0.66;0.94) 0.69(0.58;0.81) 0.000 . 0.82(0.68;0.99) 0.85(0.70;1.03) 0.018 Sweetsc 1.16(0.95;1.40) 1.32(1.09;1.59) 1.29(1.05;1.58) 0.007 1.14(0.93;1.40) 1.27(1.03;1.55) 1.23(0.98;1.54) 0.092 Sugaryc 1.25(1.03;1.53) 1.22(1.02;1.47) 1.44(1.19;1.75) 0.000 1.17(0.95;1.44) 1.10(0.90;1.34) 1.30(1.06;1.60) 0.027 Juicesd . 1.26(1.07;1.48) 1.39(1.18;1.64) 0.000 . 1.33(1.12;1.58) 1.58(1.32;1.88) 0.000 Caloric Drinksd . 0.83(0.70;0.98) 0.98(0.83;1.15) 0.520 . 0.78(0.65;0.93) 0.84(0.70;1.00) 0.031 Convenience Foodc 0.98(0.81;1.18) 1.11(0.92;1.34) 1.10(0.90;1.34) 0.212 0.87(0.72;1.07) 0.93(0.76;1.15) 0.84(0.67;1.05) 0.170 a Adjusted by sex, age, education, BMI, family history of colorectal cancer, physical activity, smoking status, caloric intake and alcohol intake as fixed effects and province of residence as a random effect. b Adjusted by sex, age, education, BMI, family history of colorectal cancer, physical activity, smoking status, caloric intake, alcohol intake and food group intake as fixed effects and province of residence as a random effect. c Reference intake is first quartile. d Reference intake is first + second quartile due to the more uniform distribution of data. 28 Table S2: Association between colorectal cancer incidence and the scores of adherence to Western, 1 Prudent and Mediterranean dietary patterns excluding in situ cases. 2 ALL MALE FEMALE n=4729 n=2662 n=2067 Co/Ca OR(95%CI) Co/Ca OR(95%CI) Co/Ca OR(95%CI) p-het WESTERN Q1 772/285 1 335/157 1 437/128 1 Q2 824/375 1.26 (1.03;1.54) 405/221 1.14 (0.87;1.49) 419/154 1.42 (1.06;1.89) Q3 831/394 1.43 (1.17;1.76) 449/255 1.31 (1.00;1.71) 382/139 1.59 (1.18;2.15) Q4 813/435 1.48 (1.18;1.85) 511/329 1.40 (1.06;1.84) 302/106 1.54 (1.10;2.15) p-trend <0.001 0.010 0.005 1SD-increase 1.19 (1.09;1.29) 1.19 (1.08;1.32) 1.18 (1.05;1.33) 0.865 PRUDENT Q1 783/390 1 485/286 1 298/104 1 Q2 812/351 0.85 (0.70;1.02) 431/229 0.82 (0.64;1.03) 381/122 0.92 (0.67;1.27) Q3 826/378 0.98 (0.81;1.19) 411/234 0.92 (0.72;1.17) 415/144 1.10 (0.81;1.51) Q4 819/370 0.90 (0.73;1.12) 373/213 0.84 (0.65;1.09) 446/157 1.02 (0.74;1.41) p-trend 0.659 0.322 0.606 1SD-increase 0.96 (0.89;1.05) 0.94 (0.86;1.04) 1.01 (0.89;1.15) 0.347 MEDITERRANEAN Q1 796/346 1 398/198 1 398/148 1 Q2 821/375 0.91 (0.75;1.11) 390/227 0.98 (0.76;1.27) 431/148 0.84 (0.63;1.12) Q3 816/348 0.71 (0.58;0.87) 425/215 0.71 (0.55;0.93) 391/133 0.72 (0.54;0.97) Q4 807/420 0.65 (0.52;0.80) 487/322 0.71 (0.55;0.92) 320/98 0.55 (0.40;0.76) p-trend <0.001 0.002 <0.001 1SD-increase 0.87 (0.80;0.94) 0.87 (0.79;0.96) 0.85 (0.75;0.96) 0.725 3 Table S3. Adjusted odds ratios for the association between proximal colon, distal colon and rectal 4 cancer incidence and scores of adherence to Western, Prudent and Mediterranean diet excluding in 5 situ cases. 6 Proximal Distal Rectal n=447 n=487 n=546 Co Ca OR(95%CI) Ca OR(95%CI) Ca OR(95%CI) p-het WESTERN Q1 772 105 1 82 1 96 1 Q2 824 110 1.01 (0.75;1.37) 133 1.61 (1.19;2.19) 131 1.26 (0.94;1.70) Q3 831 109 1.09 (0.80;1.49) 125 1.66 (1.21;2.28) 157 1.62 (1.20;2.18) Q4 813 123 1.16 (0.83;1.63) 147 1.99 (1.41;2.80) 162 1.44 (1.04;2.00) p-trend 0.325 <0.001 0.013 1SD-increase 1.07 (0.94;1.22) 1.28 (1.13;1.45) 1.22 (1.08;1.38) 0.085 PRUDENT Q1 783 111 1 129 1 149 1 Q2 812 110 0.91 (0.68;1.22) 118 0.86 (0.65;1.15) 121 0.77 (0.58;1.01) Q3 826 115 1.02 (0.75;1.37) 114 0.91 (0.68;1.22) 146 0.99 (0.75;1.30) Q4 819 111 0.91 (0.66;1.27) 126 1.01 (0.74;1.38) 130 0.80 (0.59;1.08) p-trend 0.777 0.899 0.409 1SD-increase 0.99 (0.87;1.12) 0.98 (0.87;1.11) 0.92 (0.83;1.03) 0.613 MEDITERRANEAN Q1 796 97 1 115 1 132 1 Q2 821 111 0.93 (0.69;1.26) 128 0.96 (0.72;1.27) 130 0.84 (0.64;1.11) Q3 816 106 0.75 (0.55;1.03) 113 0.73 (0.54;0.98) 129 0.67 (0.51;0.90) Q4 807 133 0.71 (0.51;0.99) 131 0.66 (0.48;0.91) 155 0.58 (0.43;0.79) p-trend 0.019 0.003 <0.001 1SD-increase 0.89 (0.79;1.01) 0.88 (0.78;0.99) 0.83 (0.74;0.93) 0.596 7 8 9 10 29 Table 4. Adjusted odds ratios for the association between proximal colon, distal colon and rectal cancer incidence and scores of adherence to Western, Prudent and Mediterranean dietary patterns. Proximal Colon Distal Colon Rectum n=457 n=503 n=560 Co Ca aORb (95%CI) Ca aORb (95%CI) Ca aORb (95%CI) p-het WESTERN Quartiles Q1 772 108 1 84 1 98 1 Q2 824 111 1.00 (0.75;1.35) 141 1.70 (1.26;2.29) 137 1.30 (0.97;1.74) Q3 831 110 1.07 (0.79;1.46) 128 1.67 (1.22;2.29) 159 1.60 (1.19;2.15) Q4 813 128 1.19 (0.85;1.66) 150 2.02 (1.44;2.84) 166 1.46 (1.05;2.01) p-trend 0.275 <0.001 0.013 1SD-increase 1.07 (0.95;1.22) 1.28 (1.13;1.45) 1.23 (1.09;1.38) 0.087 PAFc% 7% (-12%;25%) 40% (21%;60%) 27% (11%;44%) PRUDENT Quartiles Q1 783 114 1 132 1 151 1 Q2 811 113 0.92 (0.69;1.24) 123 0.91 (0.69;1.19) 124 0.79 (0.60;1.03) Q3 827 117 1.01 (0.75;1.37) 118 0.94 (0.71;1.26) 151 1.02 (0.78;1.33) Q4 819 113 0.92 (0.67;1.28) 130 1.06 (0.78;1.44) 134 0.83 (0.62;1.12) p-trend 0.798 0.680 0.545 1SD-increase 0.98 (0.87;1.11) 1.00 (0.88;1.12) 0.94 (0.84;1.05) 0.686 PAFc% 4% (-15%;24%) -8% (-28%;12%) 9% (-9%;28%) MEDITERRANEAN Quartiles Q1 796 100 1 124 1 133 1 Q2 821 113 0.92 (0.68;1.24) 131 0.92 (0.70;1.22) 136 0.87 (0.66;1.15) Q3 815 109 0.75 (0.55;1.03) 115 0.71 (0.53;0.95) 133 0.70 (0.53;0.93) Q4 808 135 0.70 (0.51;0.97) 133 0.65 (0.48;0.89) 158 0.60 (0.45;0.81) p-trend 0.017 0.002 <0.001 1SD-increase 0.89 (0.78;1.00) 0.88 (0.78;0.99) 0.84 (0.75;0.94) 0.746 PAFc% 16% (-3%;34%) 20% (3%;38%) 24% (9%;38%) a Co: Controls; Ca: Cases; SD: Standard Deviation. b Odds ratio of colon and rectal cancer associated to the adherence to the Western, Prudent and Mediterranean diet patterns adjusted by sex, age, education, BMI, family history of colorectal cancer, physical activity, smoking status, caloric intake and alcohol intake and province of residence as fixed effects. c PAF= Population attributable fraction. Proportion of colorectal cancer cases that could be prevented if all participants were in the most beneficial category of adherence to each pattern (Q1 for Western and Q4 for Prudent and Mediterranean) 30                 11 2 2 3 3 4 4 1 2 41 2 3 3 4 1 1 1 1 1 [ 1 1 1 1 ] PF=Proportion of population in the specific exposure c 100 ategory OR= Odds ratio for the especific Q Q Q Q Q Q Q Q Q Q Q Q Q Q QQ PF OR PF OR PF OR PF OR PF OR PF OR P PAF F OR PF OR                          exposure category