Plasma phospholipid fatty acid concentrations and ...

Download PDF
4 downloads 0 Views 138KB Size Report
Comment
Oct 12, 2011 - within the European Prospective Investigation into Cancer and. Nutrition–Europe Gastric Cancer (EPIC-EURGAST). Design: Fatty acids were ...
Plasma phospholipid fatty acid concentrations and risk of gastric adenocarcinomas in the European Prospective Investigation into Cancer and Nutrition (EPIC-EURGAST)1–3 Ve´ronique Chaje`s, Mazda Jenab, Isabelle Romieu, Pietro Ferrari, Christina C Dahm, Kim Overvad, Rikke Egeberg, Anne Tjønneland, Franc¸oise Clavel-Chapelon, Marie-Christine Boutron-Ruault, Pierre Engel, Birgit Teucher, Rudolf Kaaks, Anna Floegel, Heiner Boeing, Antonia Trichopoulou, Vardis Dilis, Tina Karapetyan, Amalia Mattiello, Rosario Tumino, Sara Grioni, Domenico Palli, Paolo Vineis, H Bas Bueno-de-Mesquita, Mattijs E Numans, Petra HM Peeters, Eiliv Lund, Carmen Navarro, Jose Ramo´n Quiro´s, Emilio Sa´nchez-Cantalejo, Aurelio Barricarte Gurrea, Miren Dorronsoro, Sara Regne´r, Emily Sonestedt, Elisabet Wirfa¨lt, Kay-Tee Khaw, Nick Wareham, Naomi E Allen, Francesca L Crowe, Sabina Rinaldi, Nadia Slimani, Fatima Carneiro, Elio Riboli, and Carlos A Gonza´lez ABSTRACT Background: Epidemiologic data suggest that diet is a risk factor in the etiology of gastric cancer. However, the role of dietary fatty acids, a modifiable risk factor, remains relatively unexplored. Objective: The objective of this study was to determine the association of plasma phospholipid fatty acid concentrations, as biomarkers of exogenous and endogenously derived fatty acids, with the risk of gastric adenocarcinoma in a case-control study nested within the European Prospective Investigation into Cancer and Nutrition–Europe Gastric Cancer (EPIC-EURGAST). Design: Fatty acids were measured by gas chromatography in prediagnostic plasma phospholipids from 238 cases matched to 626 controls by age, sex, study center, and date of blood donation. Conditional logistic regression models adjusted for Helicobacter pylori infection status, BMI, smoking, physical activity, education, and energy intake were used to estimate relative cancer risks. Results: Positive risk associations for gastric cancer were observed in the highest compared with the lowest quartiles of plasma oleic acid (OR: 1.72; 95% CI: 1.01, 2.94), di-homo-c-linolenic acid (OR: 1.92; 95% CI: 1.10, 3.35), a-linolenic acid (OR: 3.20; 95% CI: 1.70, 6.06), and the ratio of MUFAs to saturated fatty acids, as an indicator of stearoyl-CoA desaturase-1 enzyme activity (OR: 1.40; 95% CI: 0.81, 2.43). An inverse risk association was observed with the ratio of linoleic to a-linolenic acid (OR: 0.37; 95% CI: 0.20, 0.66). Conclusion: These data suggest that a specific prediagnostic plasma phospholipid fatty acid profile, characterized mainly by high concentrations of oleic acid, a-linolenic acid, and di-homo-c-linolenic acid, which presumably reflect both a complex dietary pattern and altered fatty acid metabolism, may be related to increased gastric cancer risk. Am J Clin Nutr 2011;94:1304–13. INTRODUCTION

GC4 presents wide international variation in incidence rates (1), which indicates that environmental and lifestyle factors, particularly diet, likely play an important role in its etiology (2). Previous epidemiologic studies have shown that high intakes of fruit and vegetables are negatively associated with GC risk

1304

(3–6), with stronger support for this association coming from case-control studies than from cohort studies (7). In contrast, a high intake of salt (5, 8) and of meats and meat products (3, 4, 9–11) has usually been linked to increased risk. Endogenous formation of nitroso compounds may account for the positive association reported between red and processed meat consumption and noncardia GC risk (12). Previous research relating dietary fat and different subtypes of dietary fat, as major dietary components, to GC has been relatively scarce, especially in prospective settings. High intakes of saturated (13, 14) and monounsaturated (13) fat have been associated with an increased risk of GC in case-control studies, 1 From the Nutrition and Metabolism, International Agency for Research on Cancer, Lyon, France (VC, MJ, IR, PF, SR, and NS); the Department of Cardiology, Aarhus University Hospital, Aalborg, Denmark (CCD); the Department of Epidemiology, School of Public Health, Aarhus University, Aarhus, Denmark (CCD and KO); the Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen, Denmark (RE and AT); UMR 1018, Nutrition, Hormones et Sante´ des Femmes, Centre de Recherche en Epide´miologie et Sante´ des Populations, Hoˆpital Paul Brousse, Villejuif, France (VC, FC-C, M-CB-R, and PE); the Division of Cancer Epidemiology, German Cancer Research Centre, Heidelberg, Germany (BT and RK); the Department of Epidemiology, German Institute of Human Nutrition PotsdamRehbruecke, Nuthetal, Germany (AF and HB); the WHO Collaborating Centre for Food and Nutrition Policies, Department of Hygiene, Epidemiology and Medical Statistics, University of Athens Medical School and Hellenic Health Foundation, Athens, Greece (AT and VD); the Hellenic Health Foundation, Athens, Greece (TK); the Department of Clinical and Experimental Medicine, Federico II University, Naples, Italy (AM); the Cancer Registry and Histopathology Unit, “Civile-MP Arezzo” Hospital, Ragusa, Italy (RT); the Nutritional Epidemiology Unit, IRCCS Istituto Nazionale Tumori, Milan, Italy (SG); the Molecular and Nutritional Epidemiology Unit, Cancer Research and Prevention Institute, ISPO, Florence, Italy (DP); The Servizio di Epidemiologia dei Tumori, Universita` di Torino and CPO-Piemonte, Turin, Italy (PV); the National Institute for Public Health and the Environment, Bilthoven, Netherlands, and the Department of Gastroenterology and Hepatology, University Medical Centre Utrecht, Utrecht, Netherlands (HBB-d-M); the Julius Center, University Medical Center Utrecht, Utrecht, Netherlands (MEN and PHMP); the Institute of Community Medicine, University of Tromsø, Tromsø, Norway (EL); the Institute of Molecular Pathology and Immunology of the University of Porto and Medical Faculty of Porto, Porto, Portugal (FC); the Epidemiology Department,

Am J Clin Nutr 2011;94:1304–13. Printed in USA. Ó 2011 American Society for Nutrition

PLASMA FATTY ACIDS AND GASTRIC CANCER RISK

whereas one case-control study reported a decreased risk of GC associated with a high intake of linoleic acid (15). Estimation of dietary fatty acid intakes is prone to many measurement errors. The conversion of food items into their fatty acid content is exceptionally complex for numerous reasons, including the variation in fatty acid composition within the same food according to cooking methods and industry supply. In contrast, biomarkers of dietary fatty acids, such as adipose tissue triglyceride and plasma or erythrocyte phospholipid fatty acid concentrations, offer objective, quantitative measures of bioavailable amounts of these nutrients irrespective of the source and quality of food, particularly for fatty acids that are not endogenously synthesized (16, 17). Thus, the use of biological markers of fat intake and fatty acid metabolism is a more meaningful approach for investigating the association between exogenous and endogenously produced fatty acids and cancer risk.

Regional Health Authority, Murcia, Spain, and CIBER de Epidemiologı´a y Salud Pu´blica, Spain (CN); the Unit of Nutrition, Environment and Cancer, Cancer Epidemiology Research Programme, Catalan Institute of Oncology, Barcelona, Spain (CAG); the Public Health and Health Planning Directorate, Asturias, Spain (JRQ); the Andalusian School of Public Health, Granada, Spain, and CIBER de Epidemiologı´a y Salud Pu´blica, Spain (ES-C); the Navarre Public Health Institute, Pamplona, Spain, and CIBER de Epidemiologı´a y Salud Pu´blica, Spain (ABG); the Public Health Division of Gipuzkoa, Basque Regional Health Department, San Sebastian, Spain, and CIBER de Epidemiologı´a y Salud Pu´blica, Barcelona, Spain (MD); the Surgical Clinic, Department of Clinical Sciences Malmo¨, Scania University Hospital, Lund University, Malmo¨, Sweden (SG); the Department of Clinical Sciences in Malmo¨, Nutrition Epidemiology, Lund University, Malmo¨, Sweden (ES and EW); the University of Cambridge Scholl of Clinical Medicine, Cambridge, United Kingdom (K-TK); the MRC Epidemiology Unit, Cambridge, United Kingdom (NW); the Cancer Epidemiology Unit, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, United Kingdom (NEA and FLC); the Department of Epidemiology and Public Health, Faculty of Medicine, Imperial College London, London, United Kingdom (ER). 2 Specific study results of the nested case-control study within EPIC (EUR-GAST) were obtained with financial support from the FP5 of the European Commission (QLG1-CT-2001-01049), Fundacio´ “LaCaixa” (exp BM06-130-0), and the Health Research Fund of the Spanish Ministry of Health (exp PI070130 and PI081420). The coordination of EPIC was financially supported by the European Commission (DG-SANCO) and the International Agency for Research on Cancer. The national cohorts are supported by Danish Cancer Society (Denmark); Ligue Contre le Cancer, Socie´te´ 3M, Mutuelle Ge´ne´rale de l’Education Nationale, Institut National de la Sante´ et de la Recherche Medicale (France); Deutsche Krebshilfe, Deutsches Krebsforschungszentrum and Federal Ministry of Education and Research (Germany); Hellenic Ministry of Health and Social Solidarity, the Stavros Niarchos Foundation and Hellenic Health Foundation (Greece); Italian Association for Research on Cancer and National Research Council (Italy); Dutch Ministry of Public Health, Welfare and Sports, Netherlands Cancer Registry, LK Research Funds, Dutch Prevention Funds, Dutch ZON (Zorg Onderzoek, Nederland), Statistics Netherlands; Norwegian Cancer Society; Health Research Fund, Regional Governments of Andalucı´a, Asturias, Basque Country, Murcia and Navarra, ISCIII RETIC (RD06/0020) (Spain); Swedish Cancer Society, Swedish Scientific Council and Regional Government of Ska˚ne and Va¨sterbotten (Sweden); Cancer Research UK, Medical Research Council UK (United Kingdom). 3 Address correspondence and reprint requests to V Chaje`s, Nutrition and Metabolism, International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France. E-mail: [email protected]. 4 Abbreviations used: DI, desaturation index; EPIC-EURGAST, European Prospective Investigation into Cancer and Nutrition–Europe Gastric Cancer; GC, gastric cancer; IARC, International Agency for Research on Cancer; PGA, Pepsinogen A; SCAG, severe chronic atrophic gastritis; SCD-1, stearoyl-CoA desaturase. Received October 11, 2010. Accepted for publication July 11, 2011. First published online October 12, 2011; doi: 10.3945/ajcn.110.005892.

1305

We determined the association of prediagnostic plasma phospholipid fatty acid concentrations with GC risk in a casecontrol study nested within EPIC—a large cohort of .520,000 subjects from 10 European countries (18, 19). In addition, because GC risk factors may differ on the basis of histologic subtype or the anatomic localization within the stomach (20), the effects of plasma fatty acids by GC subtype and subsites were also considered, as were Helicobacter pylori infection status and presence of SCAG status.

SUBJECTS AND METHODS

Details on the study population, blood sample collection procedures, follow-up methods, ascertainment of vital status, selection of study subjects, diet and lifestyle questions, nested case-control study design, and laboratory methods for H. pylori analysis are detailed elsewhere (9, 21, 22).

Study subjects, data and biosample collection, and follow-up The EPIC cohort consists of 23 subcohorts in 10 European countries (Denmark, France, Greece, Germany, Italy, the Netherlands, Norway, Spain, Sweden, and United Kingdom), with a total of 521,468 subjects enrolled. The French and Norwegian cohorts, and the Naples center of Italy and the Utrecht center in the Netherlands, consist of women only. Usual diet over the previous 12 mo was measured at recruitment by using validated country-specific diet questionnaires. Values for total energy intake and macro- and micronutrients were computed by using country-specific food-composition tables. Standardized lifestyle and personal questionnaires, various anthropometric measures, and blood samples were collected from most participants (1992– 1998). Eligible participants gave written informed consent. Approval for this study was obtained from the ethical review boards of the IARC and from all local participating centers. At each center, blood samples were drawn from participants and stored at 5–10°C, protected from light, and transported to local laboratories to be processed and portioned into aliquots (19, 23). The only exceptions were the Oxford center, where blood samples were collected from a network of general practitioners and transported to a central laboratory by post, and centers in Sweden and Denmark, where blood was portioned into aliquots within 1 h of being drawn. In all countries, blood was separated into fractions (serum, plasma, red blood cells, and buffy coat). In each country, except Denmark and Sweden, each fraction was placed into straws, which were heat-sealed and stored in liquid nitrogen (2196°C). In Denmark, blood fraction aliquots of 1.0 mL were stored locally in Nunc tubes at –150°C under nitrogen vapor. In Sweden, samples were stored in freezers at –80°C. Follow-up is based on population cancer registries (Denmark, Italy, the Netherlands, Norway, Spain, Sweden, and United Kingdom) and other methods, such as health insurance records, pathology registries and active contact of study subjects (France, Germany, and Greece). The follow-up period for the current study was for data reports received at the IARC to the end of October 2002, whichrepresented complete follow-upuntil either December 2000 or December 2001 for all centers using cancer registry data and until 2002 for France, Germany, and Greece. Subjects from

1306

CHAJE´S ET AL

Norway were not included in the current analysis because of the low number of GC cases and short period of follow-up. To reduce the effects of extreme intakes, subjects in the top and bottom 1% of the ratio of energy intake to estimated energy requirement (calculated from age, sex, height, and body weight) and subjects with previous cancer diagnoses or with missing dietary data were excluded (18 cases), as were 138 prevalent GC cases.

lysates of the CCUG H. pylori strain. PGA was assayed in plasma samples by using a commercial microplate-based quantitative ELISA kit (Biohit). SCAG was serologically defined as a circulating PGA concentration ,22 lg/L (21). Statistical methods

Plasma phospholipid fatty acids were analyzed in the same laboratory at the IARC by using a method that was previously described (24). Briefly, lipids were extracted from 200 lL plasma, phospholipids were purified by solid-phase extraction, and fatty acid methyl esters were formed by transmethylation of the phospholipids and analyzed by gas chromatography on a 30-m polar column. Cases and controls were run in the same batch. One quality-control sample originating from a standard plasma pool, located in the middle and end of the batch, was injected twice within each batch. CVs for the major fatty acids ranged from 0.93% for large peaks to 12.75% for the smallest peaks. The relative amount of each fatty acid, expressed as a percentage of total fatty acids, was quantified by integrating the area under the peak and dividing the result by the total area. Values for the individual fatty acids were used to calculate the percentage of main fatty acid groupings: SFAs, cis MUFAs, n26 PUFAs, and n23 PUFA. In addition, we also determined the DIs as the ratios of MUFAs to SFAs (DI n29; an indicator of activity of the ratelimiting enzyme D9-desaturase or SCD-1, which transforms palmitic and stearic acids into the MUFAs palmitoleic and oleic acids), c-linolenic acid to linoleic acid (DI n26; an indicator of activity of D6-desaturase), and arachidonic acid to di-homoc-linolenic acid (DI n25; an indicator of activity of D5-desaturase) (Figure 1). The ratios of n26 to n23 PUFAs and of linoleic acid to a-linolenic acid indicate the activities of the 2 parallel fatty acid metabolic pathways that compete for the desaturase and elongase enzymes.

Differences in mean fatty acids between cases and controls were tested by paired t tests of the log-transformed value in each casecontrol set. ORs and 95% CIs for GC risk in relation to each plasma individual fatty acid and fatty acid groupings were calculated by conditional logistic regression (version 9; SAS statistical software), stratified by the case-control set. Plasma phospholipid fatty acids were divided into quartiles based on the distribution among controls. Multivariate analyses were run with control for potential confounders, including H. pylori infection status (yes or no); BMI (in quartiles); status, duration, and intensity of smoking (never smokers, ex-smokers who smoked for ,10 y, ex-smokers who smoked for 10 y, smokers who smoke ,15 cigarettes/d, smokers who smoke between 15 and 25 cigarettes/d, smokers who smoke 25 cigarettes/d, or missing); physical activity (at work and leisure expressed as metabolic equivalent tasks); level of schooling (unspecified/none/primary school, technical/professional school, secondary school, or university degree; proxy variable for socioeconomic status); and energy intake (in quartiles). The effect of alcohol or meat intake was also examined. Because it did not substantially change the risk estimates, it was not included in the final model. For all models, linear trend tests were determined by using a score variable with values from 1 to 4, consistent with the quartile grouping. Subgroup analyses were performed by unconditional logistic regression, including matching variables in the model, according to sex, European region [north-central Europe (Sweden, Denmark, Germany, the Netherlands, United Kingdom, and France) compared with southern Europe (Italy, Greece, and Spain], SCAG status [yes (PGA  22 lg/L) compared with no (PGA . 22 lg/L)], H. pylori status (infection or no infection), and time from blood donation to cancer diagnosis (,2 y compared with 2 y). Tests for heterogeneity were performed by using chi-square tests. Because only 20% of cases and 8% of controls had SCAG, which resulted in small numbers of subjects in each quartile in this subgroup analyses, we performed an exact test to get unbiased estimates of the P value to evaluate relative risks in subgroups of SCAG. This allows a more accurate estimation of model parameters, CIs, and associated P values when the sample size is limited. We also investigated associations with different outcome types: anatomic subsite (cardia compared with noncardia) and histologic subtype (diffuse compared with intestinal). GC that could not be classified by anatomical subsite or histologic subtype were excluded in those subanalyses. Tests for heterogeneity were also performed by using chi-square tests.

Laboratory assay, H. pylori infection, and SCAG status

RESULTS

Nested case-control analysis and selection of study subjects A total of 238 gastric cancer cases (n = 133 men and 105 women) with available blood samples were identified and matched to 626 controls (n = 333 men and 293 women). For each identified GC case, incidence density sampling was used to identify control subjects (1:4) matched by sex, age group (6 2.5 y), study center, and date of blood sample collection (6 45 d). GC cases in the nested case-control design were also subdivided by anatomic subsite (tumors originating from the gastric cardia, n = 69; noncardial tumors, n = 124; and tumors from unknown sites, n = 45) and by histologic subtype (diffuse tumors, n = 91; intestinal tumors, n = 94). The remaining cases (n = 53) were of unknown or mixed histologic types. Analysis of plasma phospholipid fatty acids

The method used to determine H. pylori infection status was detailed elsewhere (21). The analysis for H. pylori infection status (quantification of anti-H. pylori and anti-CagA IgG antibodies) were assessed for 238 cases and their 626 matched controls in our study. Quantification of anti-H. pylori antibodies in plasma samples was done by ELISA using incubation with

Description of the study population Baseline characteristics and a description of the study population are shown in Table 1. A total of 238 GC cases and 626 matched controls with available blood samples were identified. On average, GC cases had 3.2 y between blood donation and the

PLASMA FATTY ACIDS AND GASTRIC CANCER RISK

1307

FIGURE 1. Fatty acid synthesis pathway. DPA, docosapentaenoic acid; PGE, prostaglandin E; PGG, prostaglandin G; PGH, prostaglandin H; SCD-1, stearoyl-CoA desaturase.

time of diagnosis and had a higher percentage of H. pylori and SCAG positivity than did controls. Mean plasma phospholipid fatty acid concentrations in case and control subjects in the EPIC-EURGAST Study The means (6SDs) of the percentage of plasma individual fatty acids and groupings for case and control subjects are shown in Table 2. In both cases and controls, SFA represented ;40% of total fatty acids, MUFAs ;12.5%, n26 PUFAs ;39%, and n23 PUFAs ;7%; trans MUFAs quantified are mainly represented by elaidic acid. For plasma individual fatty acids, the mean concentration of oleicacid was significantly higherincases thanin controls.For fatty acid groupings, only the mean total plasma MUFA concentration was significantly higher in cases than in controls (Table 2). Plasma phospholipid fatty acid concentrations and gastric cancer risk in the EPIC-EURGAST study The ORs for GCs according to quartiles of plasma total and individual fatty acids and the ratios of fatty acids are shown in

Table 3. No significant associations were apparent for any category of SFA, n26 PUFA, or n23 PUFA. However, for MUFA, a dose response increase in risk was observed, which was statistically significant at the highest quartile of plasma concentration (OR top compared with bottom quartile: 1.75; 95% CI: 1.04, 2.95; P-trend = 0.017). For individual MUFAs, an increased risk of GC was associated with an increasing concentration of oleic acid (OR of top compared with bottom quartile: 1.72; 95% CI: 1.01, 2.94; P-trend = 0.047). Although GC risk appeared to be higher with increasing concentrations of elaidic acid, the association was not statistically significant (OR top compared with bottom quartile: 1.63; 95% CI: 0.89, 2.97; P-trend = 0.122). Of the n26 PUFAs, a positive association was found between di-homo-c-linolenic acid and GC risk (OR top compared with bottom quartile: 1.92; 95% CI: 1.10, 3.35; P-trend = 0.03). Of the n23 PUFAs, a positive association was found between a-linolenic acid and GC risk (OR top compared with bottom quartile: 3.20, 95% CI: 1.70, 6.06; P-trend = 0.001). Longchain n23 PUFAs were not significantly associated with GC risk. Considering the ratios of fatty acids, a positive association was found between GC risk and the ratio of MUFA to SFA (DI n29)

CHAJE´S ET AL

1308

TABLE 1 Baseline characteristics of the study population of GC cases and controls in the EPIC-EURGAST Study1 GC

2

Age at recruitment Age at diagnosis2 H. pylori–positive subjects [n (%)] Subjects with SCAG [n (%)] BMI (kg/m2)2 Male subjects [n (%)] Female subjects [n (%)] Smoking status, duration, and intensity [n (%)] Never smokers Smokers, ,15 cigarettes/d Smokers, 15 to ,25 cigarettes/d Smokers, 25 cigarettes/d Ex-smokers, duration of smoking ,10 y Ex-smokers, duration of smoking 10 y Ex-smokers, missing duration of smoking Subjects with missing smoking status Mean energy intake (kcal/d)3 Median alcohol intake (g/d)3 Mean leisure and physical activity (MET-h/wk)3,4 Secondary school education or higher [n (%)] Grouping by anatomic subsite [n (%)] Cardial Noncardial Unknown or mixed subsite Grouping by histologic subsite [n (%)] Diffuse Intestinal Unknown or mixed subtype

Cases (n = 238)

Controls (n = 626)

59.2 (43.3–72.1) 62.5 (44.6–74.4) 199 (83.6) 46 (19.5) 26.2 (20.5–32.8) 133 (55.9) 105(44.1)

59.3 (43.7–72.1) — 416 (66.5) 51 (8.2) 26.6 (20.9–34.2) 333 (53.2) 293 (46.8)

81 24 28 11 10 70 5 9 2174 6.3 81.4 69

286 55 42 13 26 169 10 25 2184 6.7 85.6 201

(34.0) (10.1) (11.8) (4.6) (4.2) (29.4) (2.1) (3.8) (736–4642) (0–152.0) (0–256.5) (29.0)

(45.7) (8.8) (6.7) (2.1) (4.2) (27) (1.6) (4.0) (600–7244) (0–442.7) (0–296.4) (32.1)

69 (29.0) 124 (52.1) 45 (18.9)

— — —

91 (38.2) 94 (39.5) 53 (22.3)

— — —

1 Distribution of cases/controls by EPIC center: Greece, 12/27; Spain, 25/72; Italy, 44/149; France, 3/10; Germany, 30/ 87; Netherlands, 19/60; Sweden, 56/109; Denmark, 21/34; United Kingdom, 28/78. EPIC-EURGAST, European Prospective Investigation into Cancer and Nutrition–Europe Gastric Cancer; GC, gastric cancer; MET-h, metabolic equivalent task hours; SCAG, severe chronic atrophic gastritis. 2 Values are means; 5th–95th percentile range in parentheses. 3 Range in parentheses. 4 Combined recreational and household activity.

(OR top compared with bottom quartile: 1.40; 95% CI: 0.81, 2.43; P-trend = 0.053). No significant association was found between GC risk and the ratio of c-linolenic acid to linoleic acid (DI n26) (OR top compared with bottom quartile: 1.56; 95% CI: 0.95, 2.56; P-trend = 0.235) nor with the ratio of arachidonic acid to di-homo-c-linolenic acid (DI n25) (OR top compared with bottom quartile: 0.67; 95% CI: 0.42, 1.06; P-trend = 0.109). Finally, a negative association was found between the ratio of linoleic acid to a-linolenic acid and GC risk (OR top compared with bottom quartile: 0.37; 95% CI: 0.20, 0.66; P-trend = 0.002) (Table 3). No significant GC risk associations were found for any of the individual plasma fatty acids and groupings by anatomic subsite (P-heterogeneity: SFA, 0.12; MUFA, 0.49; n26 PUFA, 0.71; n23 PUFA, 0.68), histologic subtype (P-heterogeneity: SFA, 0.78; MUFA, 0.50; n26 PUFA, 0.59; n23 PUFA, 0.95), or sex (P-heterogeneity: SFA, 0.97; MUFA, 0.91; n26 PUFA, 0.69; n23 PUFA, 0.34) (data not tabulated). Consideration of H. pylori infection status showed a positive statistically significant GC risk association in H. pylori–positive subjects (n = 199 cases and 416 controls) with MUFA concen-

tration (OR of top compared with bottom quartile: 1.95; 95% CI: 1.08, 3.52; P-trend = 0.02), although P-heterogeneity for the effect between H. pylori–positive and –negative subjects was not significant (P = 0.34). The effect for MUFA was likely due to oleic acid (OR of top compared with bottom quartile: 2.08; 95% CI: 1.13, 3.83; P-trend = 0.03). Associations between individual SFAs, n26 PUFAs, or n23 PUFAs and GC risk did not differ between H. pylori–positive and H. pylori–negative subjects (data not tabulated). Analyses by European region showed that the positive association between oleic acid and GC risk tended to be more pronounced in the north-central region than in the south, although the P value for heterogeneity was not significant (P = 0.58) (data not tabulated). The positive association between di-homo-c-linolenic acid (P-heterogeneity = 0.04) or a-linolenic acid (P-heterogeneity = 0.31) and GC risk was stronger in the south than in the north-central region (data not tabulated). Because the main source of oleic acid and a-linolenic acid in north-central Europe is meat intake, we further analyzed the associations between these specific fatty acids and GC risk separately in the north-central and southern regions, with further

1309

PLASMA FATTY ACIDS AND GASTRIC CANCER RISK TABLE 2 Mean plasma phospholipid fatty acids at baseline in cases and controls in the EPIC-EURGAST Study1 Individual plasma phospholipid fatty acids Total SFAs 15:0, Pentadecanoic acid 16:0, Palmitic acid 17:0, Heptadecanoic acid 18:0, Stearic acid Total MUFAs, cis 16:1n27, Palmitoleic acid 18:1n29, Oleic acid MUFAs, trans 18:1n29trans, Elaidic acid Total n26 PUFAs 18:2n26, Linoleic acid 18:3n26, c-linolenic acid 20:2n26, Di-homo-linoleic acid 20:3n26, Di-homo-c-linolenic acid 20:4n26, Arachidonic acid Total n23 PUFAs 18:3n23, a-Linolenic acid 20:5n23, Eicosapentaenoic acid 22:5n23, Docosapentaenoic acid 22:6n23, Docosahexaenoic acid

Cases (n = 238)

Controls (n = 626)

P value for difference2

% of total fatty acids 40.89 6 2.24 0.17 6 0.05 27.17 6 2.91 0.37 6 0.08 12.76 6 1.20 12.59 6 2.16 0.73 6 0.30 10.68 6 1.99

% of total fatty acids 40.79 6 2.13 0.16 6 0.05 26.93 6 2.83 0.37 6 0.08 12.92 6 1.22 12.38 6 2.29 0.70 6 0.26 10.50 6 2.15

0.52 0.99 0.67 0.97 0.82 0.03 0.39 0.02

6 6 6 6 6 6 6 6 6 6 6 6

0.48 0.26 0.18 0.38 0.98 0.14 0.93 0.85 0.85 0.75 0.14 0.94

0.08 39.03 23.36 0.13 0.43 4.50 10.01 7.41 0.21 1.17 1.13 4.89

6 6 6 6 6 6 6 6 6 6 6 6

0.10 3.31 3.32 0.09 0.08 1.15 2.08 2.13 0.09 0.31 0.86 1.39

0.07 39.45 23.51 0.12 0.44 4.49 10.27 7.30 0.20 1.17 1.07 4.86

0.08 3.40 3.48 0.07 0.08 1.13 2.26 2.02 0.09 0.31 0.69 1.41

1 Cases: n = 238 (total), n = 84 (southern regions), n = 154 (northern regions); control subjects: n = 626 (total), n = 258 (southern regions), n = 368 (northern regions). Southern regions: Italy, Spain, and Greece; northern regions: France, Germany, United Kingdom, Netherlands, Sweden, and Denmark. EPIC-EURGAST, European Prospective Investigation into Cancer and Nutrition–Europe Gastric Cancer. 2 Differences between cases and controls in mean concentrations of each fatty acid were tested by paired t tests of the log-transformed value in each case-control set.

adjustment for meat intake. However, this did not change the risk estimates (data not tabulated). The potential effect modification by the time to diagnosis of GC of ,2 y or .2 y showed no overall significant heterogeneity (data not tabulated). The potential effect modification by SCAG status showed no significant heterogeneity for the association between di-homo-c-linolenic acid (P = 0.28) or a-linolenic (P = 0.15) acid and GC risk (data not tabulated). DISCUSSION

In this nested case-control study based on the prospective EPIC cohort, we found evidence of increased GC risk associated with increasing concentrations of oleic acid, di-homo-c-linolenic acid, and a-linolenic acid. The ratio of MUFA to SFA, as a putative index of activity of SCD-1, was positively associated with GC risk. Grouping of individual fatty acids into their main families showed no statistically significant associations with GC risk, except for MUFA, increased concentrations of which were strongly associated with higher risk. This association was largely related to oleic acid, the main dietary MUFA. Oleic acid is not an essential fatty acid and can be either derived from the diet or formed endogenously by the desaturation of stearic acid into oleic acid via the microsomal enzyme SCD-1. Dietary MUFAs originated mainly from olive oil in southern Europe and from meat in north-central Europe (25). However, in a cross-sectional study within the EPIC cohort, the weak correlations calculated at the

individual level between plasma phospholipid oleic acid and olive oil or meat intake suggested that dietary contributors of plasma oleic acid concentrations may not be strong determinants compared with endogenous hepatic synthesis, mainly in countries with low olive intake (16). Thus, an increased risk of GC associated with a high level of oleic acid along with a high ratio of MUFAs to SFAs, as previously reported in GC (26) and breast cancer (24, 27, 28), might be mostly related to increased desaturation of stearic acid into oleic acid rather than to a diet rich in MUFAs. Diet may also have an important effect on SCD-1 activity. In EPIC, a high consumption of alcohol was associated with increased ratio of MUFA to SFA, whereas a high consumption of olive oil counteracted this effect (16, 29). These data suggested that high plasma phospholipid oleic acid, presumably reflecting both a specific complex dietary pattern and increased stearic acid desaturation via SCD-1, is related to increased risk of GC. The potential pathway underlying the association between an increased prediagnostic plasma MUFA to SFA ratio and increased risk of cancer is not known. Despite the overactivation of enzymes that synthesize SFAs (30), abundant amounts of MUFAs are found in cancer cells (31, 32). It is known that MUFAs can serve as mediators of signal transduction and cellular differentiation, and unbalanced concentrations of these mediators have also been implicated in carcinogenesis (33). The expression of SCD-1, the main SCD isoform in humans, is elevated in several human cancers and chemically induced tumors (31, 34, 35). The suppression of SCD-1 expression reduces cancer cell

CHAJE´S ET AL

1310

TABLE 3 Association of gastric cancer risk with plasma phospholipid fatty acid concentrations at baseline in the EPIC-EURGAST Study1 Quartile of fatty acid Plasma fatty acids SFAs (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Pentadecanoic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Palmitic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Heptadecanoic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Stearic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) MUFA, cis Range (%) Cases/controls (n) OR (95% CI) Palmitoleic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Oleic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Elaidic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) n26 PUFAs (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Linoleic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) c-Linolenic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Di-homo-linoleic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) Di-homo-c-linolenic acid Range (%) Cases/controls (n) OR (95% CI) Arachidonic acid (% of total fatty acids) Range (%) Cases/controls (n)

1 (reference)

2

3

4

P-trend

,39.6 16/157 1.00

39.6 to ,40.89 26/156 0.97 (0.57, 1.65)

40.89 to ,42.27 55/156 0.91 (0.52, 1.60)

42.27 71/157 1.17 (0.59, 2.34)

0.777

,0.13 63/157 1.00

0.13 to ,0.16 62/156 1.05 (0.65, 1.69)

0.16 to ,0.19 48/157 0.66 (0.38, 1.15)

0.19 65/156 0.90 (0.51, 1.58)

0.520

,25.29 58/157 1.00

25.29 to ,27.08 49/156 0.66 (0.38, 1.14)

27.08 to ,28.76 55/157 0.74 (0.42, 1.30)

28.76 76/156 0.99 (0.52, 1.89)

0.906

,0.32 61/156 1.00

0.32 to ,0.37 53/157 1.12 (0.70, 1.81)

0.37 to ,0.42 56/156 0.99 (0.62, 1.59)

0.42 68/157 1.11 (0.66, 1.87)

0.830

,12.12 75/156 1.00

12.12 to ,12.87 49/157 0.69 (0.43, 1.11)

12.87 to ,13.70 64/156 1.08 (0.68, 1.72)

13.70 50/157 1.06 (0.61, 1.84)

0.593

,10.83 46/156 1.00

10.83 to ,12.14 59/157 1.19 (0.72, 1.98)

12.14 to ,13.64 63/156 1.40 (0.84, 2.34)

13.64 70/157 1.78 (1.05, 3.00)

0.026

,0.52 51/156 1.00

0.52 to ,0.66 64/157 1.34 (0.80, 2.25)

0.66 to ,0.83 63/156 1.20 (0.70, 2.06)

0.83 60/157 1.05 (0.59, 1.87)

0.901

,9.00 47/157 1.00

9.00 to ,10.23 61/156 1.30 (0.79, 2.16)

10.23 to ,11.71 66/157 1.45 (0.87, 2.39)

11.71 64/156 1.72 (1.01, 2.94)

0.047

,0.021 50/156 1.00

0.021 to ,0.045 55/157 1.26 (0.77, 2.06)

0.045 to ,0.094 60/157 1.36 (0.80, 2.34)

0.094 73/156 1.63 (0.89, 2.97)

0.122

,37.10 60/157 1.00

37.10 to ,39.56 74/156 1.39 (0.89, 2.19)

39.56 to ,41.52 49/157 1.06 (0.64, 1.75)

41.52 55/156 1.15 (0.67, 1.97)

0.865

,20.99 54/157 1.00

20.99 to ,23.48 66/156 1.09 (0.68, 1.74)

23.48 to ,25.80 62/157 1.06 (0.65, 1.73)

25.80 56/156 0.90 (0.55, 1.48)

0.660

,0.077 61/157 1.00

0.077 to ,0.11 51/156 1.05 (0.64, 1.71)

0.11 to ,0.155 64/157 1.24(0.77, 1.99)

0.155 62/156 1.31 (0.81, 2.11)

0.214

,0.38 66/156 1.00

0.38 to ,0.43 59/157 0.96 (0.60, 1.53)

0.43 to ,0.48 58/156 1.03 (0.65, 1.64)

0.48 55/157 1.26 (0.76, 2.07)

0.372

,3.71 60/156 1.00

3.71 to ,4.39 59/157 1.32 (0.82, 2.13)

4.39 to ,5.20 59/157 1.36 (0.81, 2.28)

5.20 60/156 1.92 (1.10, 3.35)

0.030

,8.72 73/157

8.72 to ,10.25 59/156

10.25 to ,11.68 54/156

11.68 52/157 (Continued)

1311

PLASMA FATTY ACIDS AND GASTRIC CANCER RISK TABLE 3 (Continued ) Quartile of fatty acid Plasma fatty acids OR (95% CI) n23 PUFAs Range (%) Cases/controls (n) OR (95% CI) a-Linolenic acid (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) EPA (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) DPA (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) DHA (% of total fatty acids) Range (%) Cases/controls (n) OR (95% CI) MUFA/SFA (DI n29) Range (DI n29) Cases/controls (n) OR (95% CI) c-Linolenic acid/linoleic acid (DI n26) Range Cases/controls (n) OR (95% CI) Arachidonic acid/di-homo-c-linolenic acid (DI n25) Range Cases/controls (n) OR (95% CI) n26 PUFAs/n23PUFAs Range Cases/controls (n) OR (95% CI) Linoleic acid/a-linolenic acid Range Cases/controls (n) OR (95% CI)

1 (reference)

2

3

4

P-trend

1.00

0.93 (0.59, 1.48)

0.77 (0.45, 1.30)

0.86 (0.49, 1.51)

0.516

,5.90 52/156 1.00

5.90 to ,6.93 63/157 1.25 (0.77, 2.02)

6.93 to ,8.41 62/156 1.24 (0.74, 2.06)

8.41 61/157 0.96 (0.58, 1.60)

0.776

,0.13 40/157 1.00

0.13 to ,0.18 56/156 1.98 (1.14, 3.44)

0.18 to ,0.24 54/156 2.11 (1.13, 3.94)

0.24 88/157 3.20 (1.70, 6.06)

0.001

,0.62 47/157 1.00

0.62 to ,0.87 64/156 1.49 (0.91, 2.43)

0.87 to ,1.30 62/156 0.98 (0.59, 1.64)

1.30 65/157 0.95 (0.55, 1.64)

0.456

,0.96 66/157 1.00

0.96 to ,1.13 58/156 0.86 (0.53, 1.41)

1.13 to ,1.37 53/157 0.81 (0.47, 1.40)

1.37 61/156 0.88 (0.47, 1.63)

0.642

,3.83 53/156 1.00

3.83 to ,4.70 63/157 1.22 (0.76, 1.96)

4.70 to ,5.63 56/157 1.16 (0.72, 1.88)

5.63 66/156 1.09 (0.65, 1.81)

0.846

,0.26 52/156 1.00

0.26 to ,0.30 51/157 0.86 (0.51, 1.45)

0.30 to ,0.34 72/156 1.52 (0.93, 2.49)

0.34 63/157 1.40 (0.81, 2.43)

0.053

,0.0031 51/156 1.00

0.0031 to ,0.0049 72/157 1.89 (1.17, 3.07)

0.0049 to ,0.0070 56/156 1.37 (0.84, 2.24)

0.0070 59/157 1.56 (0.95, 2.56)

0.235

,1.93 67/156 1.00

1.93 to ,2.30 57/157 0.84 (0.55, 1.31)

2.30 to ,2.77 62/156 0.85 (0.54, 1.33)

2.77 52/157 0.67 (0.42, 1.06)

0.109

,4.59 69/156 1.00

4.59 to ,5.77 60/157 1.07 (0.68, 1.67)

5.77 to ,6.91 59/157 1.09 (0.68, 1.73)

6.91 50/156 0.89 (0.54, 1.46)

0.721

100.87 to ,127.04 47/157 0.56 (0.35, 0.92)

127.04 to ,170.96 58/156 0.64 (0.40, 1.03)

170.96 43/157 0.37 (0.20, 0.66)

0.002

,100.87 90/156 1.00

1

n = 238 cases and 626 controls. Values are ORs (95% CIs) derived from a model adjusted for Helicobacter pylori infection status (yes or no); BMI (in quartiles); status, duration, and intensity of smoking (never smokers, ex-smokers who smoked for ,10 y, ex-smokers who smoked for 10 y, smokers who smoke ,15 cigarettes/d, smokers who smoke between 15 and 25 cigarettes/d, smokers who smoke 25 cigarettes/d, and missing); physical activity (at work and leisure, expressed as metabolic equivalent tasks); level of schooling (unspecified, none, primary school, technical/professional school, secondary school, or university degree); and energy intake (in quartiles). DI, desaturation index; DPA, docosapentaenoic acid; EPIC-EURGAST, European Prospective Investigation into Cancer and Nutrition–Europe Gastric Cancer.

proliferation and in vitro invasiveness and dramatically impairs tumor formation and growth (36, 37). These data may suggest that endogenously synthesized MUFAs, rather than exogenous MUFAs, act as regulators of cancer cell growth (38). Another MUFA that deserves attention with respect to cancer risk is elaidic acid, the prevailing trans MUFA isomer occurring during partial hydrogenation of vegetable oils and used as ingredient in the formulation of industrial processed foods (39). Only a few studies have examined the association between industrial trans fatty acid intake and cancer risk, but most of them reported evidence of increased risks of breast cancer (27, 40),

distal colorectal cancer (41), and prostate cancer (42) associated with increasing concentrations of industrially produced trans fatty acids. Here, we also found a suggestion of a positive association between GC risk and plasma phospholipid elaidic acid concentrations. The lack of a significant association may be explained by the fact that the laboratory method used a 30-m column, which was nonspecific for the separation of cis- from trans fatty acids, hence not allowing complete separation of elaidic acid from oleic acid. The 2 PUFA subtypes may lead to production of different eicosanoid families, which have vastly different inflammatory

1312

CHAJE´S ET AL

effects in various tissues, including those of the stomach. Inflammation, particularly due to H. pylori infection, is a major gastric risk factor, and increased expression of cyclooxygenase2, key regulator of prostaglandin metabolism, has been suggested to be an early event in GC (43). The end products of cyclooxygenase-2 activity consist of a panel of prostaglandins and thromboxanes, which are critical regulators of fundamental physiologic and pathologic processes, including platelet aggregation, T cell development, inflammation, and cancer (44). Indeed, use of nonsteroidal antiinflammatory medications that inhibit cyclooxgenase-2 have been suggested to decrease GC risk (45). In the current study, a higher concentration of di-homoc-linolenic acid was associated with an increased GC risk, as previously reported in prostate cancer, particularly advanced tumors (46). The main dietary sources of di-homo-c-linolenic acid are oil seeds, such as borage, evening primrose, and black currant oils, but its blood concentrations are likely mostly due to its synthesis from the metabolism of linoleic acid via D6-desaturation and elongation (16, 29). This could explain the apparently discrepant findings regarding the association between high borage oil consumption and decreased GC risk (47). However, we found no significant association between risk of GC and the ratio of c-linolenic acid to linoleic acid, as an index of D6-desaturase activity, or with the ratio of arachidonic acid to di-homo-c-linolenic acid, as an index of D5-desaturase. The current study showed a strong association between an increased risk of GC and increasing concentrations of a-linolenic acid—the essential fatty acid of the n23 series. a-Linolenic acid is present in flaxseed oil, some vegetable and nut oils, and animal fats. Several epidemiologic studies have reported strong associations between high dietary intake or blood concentrations of a-linolenic acid and increased risks of breast (48), colorectal (49), and prostate (50–52) Cancers. In these studies, adjustment for animal or nonanimal sources did not change the cancer risk estimates for a-linolenic acid. However, a study reported decreased breast cancer risk associated with a high consumption of a-linolenic acid from fruit and vegetables and an increased risk associated with high consumption of a-linolenic acid from nut mixes and processed meat (53). Because a-linolenic acid is preferentially oxidized among PUFAs, an alternative hypothesis might be that lower oxidation of a-linolenic acid is associated with higher GC risk. Further studies are required to determine whether altered metabolism of a-linolenic acid has a mechanistic role in GC etiology or whether this fatty acid is a biomarker of a specific food pattern that may affect differentially the risk of this cancer. Even if a strength of this study is the measurement of the plasma concentration of fatty acids in a subset of cases from which blood samples were collected before cancer diagnosis, the possibility that prediagnostic SCAG may have affected absorption and metabolism of fatty acids (reverse causation), because it was the case for vitamins B-6 and riboflavin (54), cannot be ruled out, despite the absence of statistical heterogeneity by time since diagnosis and SCAG status. Finally, we cannot exclude results due to chance alone because of the high number of tests performed. In summary, findings from this study suggest that a prediagnostic plasma phospholipid fatty acid profile characterized mainly by high concentrations of oleic acid, a-linolenic acid, and di-homo-c-linolenic acid—which presumably reflects both a complex dietary pattern and impaired metabolism—is associated with an increased risk of GC.

We thank Carine Biessy for the statistical analyses and the members of the pathologist panel for their valuable work: Roger Stenling (Umea, Sweden), Johan Offerhaus (Amsterdam, Netherlands), Vicki Save (Cambridge, United Kingdom), Gabriella Nesi (Firenze, Italy), U Mahlke (Postdam, Germany), Hendrik Bla¨ker (Heildelberg, Germany), Claus Fenger (Denmark), and Dimitrious Roukos (Ioannina, Greece) for collaboration in the collection of pathologic material and Catia Moutinho (Porto, Portugal) for technical work in the preparation of the pathologic material. The authors’ responsibilities were as follows—CAG: coordinated the EPIC-EURGAST Study; VC, MJ, and CAG: wrote the manuscript, conducted the analyses, and prepared the manuscript; and VC: set up the laboratory method of analysis of plasma fatty acids and metabolism. All of the other authors contributed to and/or supervised the collection and analysis of the dietary data and the collection of blood samples in the participating centers and/or provided comments and suggestions on the intermediate and final versions of the manuscript. None of the authors had any financial or personal conflicts of interest to disclose.

REFERENCES 1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. GLOBOCAN 2008, cancer incidence and mortality Worldwide: IARC CancerBase no. 10. Lyon, France: International Agency for Research on Cancer. 2010. Available from: http://globocan.iarc.fr (cited). 2. World Cancer Research Fund. Food, nutrition, physical activity, and the prevention of cancer: a global perspective. Washington, DC: American Institute for Cancer Research, 2007. 3. Campbell PT, Sloan M, Kreiger N. Dietary patterns and risk of incident gastric adenocarcinoma. Am J Epidemiol 2008;167:295–304. 4. Navarro Silvera SA, Mayne ST, Risch H, Gammon MD, Vaughan TL, Chow WH, Dubrow R, Schoenberg JB, Stanford JL, West AB, et al. Food group intake and risk of subtypes of esophageal and gastric cancer. Int J Cancer 2008;123:852–60. 5. Tsugane S, Sasazuki S. Diet and the risk of gastric cancer: review of epidemiological evidence. Gastric Cancer 2007;10:75–83. 6. Gonza´lez CA, Pera G, Agudo A, Bueno-de-Mesquita HB, Ceroti M, Boeing H, Schulz M, Del Giudice G, Plebani M, Carneiro F, et al. Fruit and vegetable intake and the risk of stomach and oesophagus adenocarcinoma in the European Prospective Investigation into Cancer and Nutrition (EPIC-EURGAST). Int J Cancer 2006;118:2559–66. 7. Riboli E, Norat T. Epidemiologic evidence of the protective effect of fruit and vegetables on cancer risk. Am J Clin Nutr 2003;78(suppl 3): 559S–69S. 8. Tsugane S. Salt, salted food intake, and risk of gastric cancer: epidemiologic evidence. Cancer Sci 2005;96:1–6. 9. Gonza´lez CA, Jakszyn P, Pera G, Agudo A, Bingham S, Palli D, Ferrari P, Boeing H, del Giudice G, Plebani M, et al. Meat intake and risk of stomach and esophageal adenocarcinoma within the European Prospective Investigation into Cancer and Nutrition (EPIC). J Natl Cancer Inst 2006;98:345–54. 10. Larsson SC, Bergkvist L, Wolk A. Processed meat consumption, dietary nitrosamines and stomach cancer risk in a cohort of Swedish women. Int J Cancer 2006;119:915–9. 11. Van den Brandt PA, Botterweck AA, Goldbohm RA. Salt intake, cured meat consumption, refrigerator use and stomach cancer incidence: a prospective cohort study (Netherlands). Cancer Causes Control 2003; 14:427–38. 12. Jakszyn P, Bingham S, Pera G, Agudo A, Luben R, Welch A, Boeing H, Del Giudice G, Palli D, Saieva C, et al. Endogenous versus exogenous exposure to N-nitroso coumpounds and gastric cancer risk in the European Prospective Investigation into Cancer and Nutrition (EPICEURGAST) study. Carcinogenesis 2006;27:1497–501. 13. Lo´pez-Carrillo L, Lo´pez-Cervantes M, Ward MH, Bravo-Alvarado J, Ramı´rez-Espitia A. Nutrient intake and gastric cancer in Mexico. Int J Cancer 1999;83:601–5. 14. Corne´e J, Pobel D, Riboli E, Guyader M, He´mon B. A case-control study of gastric cancer and nutritional factors in Marseille, France. Eur J Epidemiol 1995;11:55–65. 15. Kaaks R, Tuyns AJ, Haelterman M, Riboli E. Nutrient intake patterns and gastric cancer risk: a case-control study in Belgium. Int J Cancer 1998;78:415–20.

PLASMA FATTY ACIDS AND GASTRIC CANCER RISK 16. Saadatian-Elahi M, Slimani N, Chaje`s V, Jenab M, Goudable J, Biessy C, Ferrari P, Byrnes G, Autier P, Peeters PH, et al. Plasma phospholipid fatty acid profiles and their association with food intakes: Results from a cross-sectional study within the European Prospective Investigation into Cancer and Nutrition (EPIC). Am J Clin Nutr 2009;89:331–46. 17. Baylin A, Campos H. The use of fatty acid biomarkers to reflect dietary intake. Curr Opin Lipidol 2006;17:22–7. 18. Bingham S, Riboli E. Diet and cancer—the European Prospective Investigation into Cancer and Nutrition. Nat Rev Cancer 2004;4:206–15. 19. Riboli E, Kaaks R. The EPIC Project: rationale and study design. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol 1997;26(suppl 1):S6–14. 20. Epplein M, Nomura AM, Hankin JH, Blaser MJ, Perez-Perez G, Stemmermann GN, Wilkens LR, Kolonel LN. Association of Helicobacter pylori infection and diet on the risk of gastric cancer: a case-control study in Hawaii. Cancer Causes Control 2008;19:869–77. 21. Palli D, Masala G, Del Guidice G, Plebani M, Basso D, Berti D, Numans ME, Ceroti M, Peeters PH, Bueno de Mesquita HB, et al. CagA+ Helicobacter pylori infection and gastric cancer risk in the EPIC-EURGAST study. Int J Cancer 2007;120:859–67. 22. Jenab M, Riboli E, Ferrari P, Friesen M, Sabate J, Norat T, Slimani N, Tjønneland A, Olsen A, Overvad K, et al. Plasma and dietary carotenoid, retinal and tocopherol levels and the risk of gastric adenocarcinomas in the European Prospective Investigation into Cancer and Nutrition. Br J Cancer 2006;95:406–15. 23. Riboli E, Hunt KJ, Slimani N, Ferrari P, Norat T, Fahey M, Charrondie`re UR, He´mon B, Casagrande C, Vignat J, et al. European Prospective Investigation into Cancer and Nutrition (EPIC): study populations and data collection. Public Health Nutr 2002;5:1113–24. 24. Chaje`s V, Hulten K, van Kappel AL, Winkvist A, Kaaks R, Hallmans G, Lenner P, Riboli E. Fatty acid composition in serum phospholipids and risk of breast cancer: an incident case-control study in Sweden. Int J Cancer 1999;83:585–90. 25. Freisling H, Fahey MT, Moskal A, Ocke´ MC, Ferrari P, Jenab M, Norat T, Naska A, Welch AA, Navarro C, et al. Region-specific nutrient intake patterns exhibit a geographical gradient within and between European countries. J Nutr 2010;140:1280–6. 26. Kuriki K, Wakai K, Matsuo K, Hiraki A, Suzuki T, Yamamura Y, Yamao K, Nakamura T, Tatematsu M, Tajima K. Gastric cancer risk and erythrocyte composition of docosahexaenoic acid with antiinflammatory effects. Cancer Epidemiol Biomarkers Prev 2007;16: 2406–15. 27. Chaje`s V, Thie´baut A, Rotival M, Gauthier E, Maillard V, Boutron-Ruault MC, Joulin V, Lenoir GM, Clavel-Chapelon F. Association between serum trans-monounsaturated fatty acids and breast cancer risk in the E3N-EPIC Study. Am J Epidemiol 2008;167:1312–20. 28. Pala V, Krogh V, Muti P, Chaje`s V, Riboli E, Micheli A, Saadatian M, Sieri S, Berrino F. Erythrocyte membrane fatty acids and subsequent breast cancer: a prospective Italian study. J Natl Cancer Inst 2001;93: 1088–95. 29. Thie´baut AC, Rotival M, Gauthier E, Lenoir G, Boutron-Ruault MC, Joulin V, Clavel-Chapelon F, Chaje`s V. Correlation between serum phospholipid fatty acids and dietary intakes assessed a few years earlier. Nutr Cancer 2009;61:500–9. 30. Swinnen JV, Vanderhoydonc F, Elagamal AA, Eelen M, Vercaeren I, Joniau S, Van Poppel H, Baert L, Goossens K, Heyns W, et al. Selective activation of the fatty acid synthesis pathway in human prostate cancer. Int J Cancer 2000;88:176–9. 31. Scaglia N, Caviglia JM, Igal RA. High stearoyl-CoA desaturase protein and activity levels in simian virus 40 transformed-human lung fibroblasts. Biochim Biophys Acta 2005;1687:141–51. 32. Bougnoux P, Chaje`s V, Lanson M, Hacene K, Body G, Couet C, Le Floch O. Prognostic significance of tumor phosphatidylcholine stearic acid level in breast carcinoma. Breast Cancer Res Treat 1992;20: 185–94. 33. Miyazaki M, Kim YC, Gray-Keller MP, Attie AD, Ntambi JM. The biosynthesis of hepatic cholesterol esters and triglycerides is impaired in mice with a disruption of the gene for stearoyl-CoA desaturase 1. J Biol Chem 2000;275:30132–8. 34. Yahagi N, Shimano H, Hasegawa K, Ohashi K, Matsuzaka T, Najima Y, Sekiya M, Tomita S, Okazaki H, Tamura Y, et al. Coordinate activation of lipogenic enzymes in hepatocellular carcinoma. Eur J Cancer 2005;41:1316–22.

1313

35. Kumar-Sinha C, Ignatoski KW, Lippman ME, Ethier SP, Chinnaiyan AM. Transcriptome analysis of HER2 reveals a molecular connection to fatty acid synthesis. Cancer Res 2003;63:132–9. 36. Scaglia N, Chisholm JW, Igal RA. Inhibition of stearoylCoA desaturase-1 inactivates acetyl-CoA carboxylase and impairs proliferation in cancer cells: role of AMPK. PLoS ONE 2009;4:e6812. 37. Scaglia N, Igal RA. Inhibition of stearoyl-CoA desaturase 1 expression in human lung adenocarcinoma cells impairs tumorigenesis. Int J Oncol 2008;33:839–50. 38. Chaje`s V, Joulin V, Clavel-Chapelon F. The fatty acid desaturation index of blood lipids, as a biomarker of hepatic stearoyl-CoA desaturase expression, is a predictive factor of breast cancer risk. Curr Opin Lipidol 2011;22:6–10. 39. Martin CA, Milinsk MC, Visentainer JV, Matsushita M, De Souza NE. Trans fatty acid-forming processes in foods: a review. An Acad Bras Cienc 2007;79:343–50. 40. Bougnoux P, Giraudeau B, Couet C. Diet, cancer, and the lipidome. Cancer Epidemiol Biomarkers Prev 2006;15:416–21. 41. Vinikoor LC, Millikan RC, Satia JA, Schroeder JC, Martin CF, Ibrahim JG, Sandler RS. Trans fatty acid consumption and its association with distal colorectal cancer in the North Carolina Colon Cancer Study II. Cancer Causes Control 2010;21:171–80. 42. Chavarro JE, Stampfer MJ, Campos H, Kurth T, Willett WC, Ma J. A prospective study of trans fatty acid levels in blood and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev 2008;17:95–101. 43. Fujimura T, Ohta T, Oyama K, Miyashita T, Miwa K. Role of cyclooxygenase-2 in the carcinogenesis of gastrointestinal tract cancers: a review and report of personal experience. World J Gastroenterol 2006;12:1336–45. 44. Smith WL, De Witt DL, Garavito RM. Cyclooxygenase: structural, cellular, and molecular biology. Annu Rev Biochem 2000;69: 145–82. 45. Wang WH, Huang JQ, Zheng GF, Lam SK, Karlberg J, Wong BC. Non-steroidal anti-inflammatory drug use and the risk of gastric cancer: a systematic review and meta-analysis. J Natl Cancer Inst 2003;95: 1784–91. 46. Chavarro JE, Stampfer MJ, Li H, Campos H, Kurth T, Ma J. A prospective study of polyunsaturated fatty acid levels in blood and prostate cancer risk. Cancer Epidemiol Biomarkers Prev 2007;16:1364–70. 47. Gonza´lez CA, Sanz JM, Marcos G, Pita S, Brullet E, Saigi E, Badia A, Agudo A, Riboli E. Borage consumption as a possible gastric cancer protective factor. Cancer Epidemiol Biomarkers Prev 1993;2:157–8. 48. De Ste´fani E, Deneo-Pellegrini H, Mendilaharsu M, Ronco A. Essential fatty acids and breast cancer: a case-control study in Uruguay. Int J Cancer 1998;76:491–4. 49. Daniel CR, McCullough ML, Patel RC, Jacobs EJ, Flanders WD, Thun MJ, Calle EE. Dietary intake of x-6 and x-3 fatty acids and risk of colorectal cancer in a prospective cohort of U.S. men and women. Cancer Epidemiol Biomarkers Prev 2009;18:516–25. 50. Crowe FL, Allen NE, Appleby PN, Overvad K, Aardestrup IV, Johnsen NF, Tjønneland A, Linseisen J, Kaaks R, Boeing H, et al. Fatty acid composition of plasma phospholipids and risk of prostate cancer in a case-control analysis nested within the European Prospective Investigation into Cancer and Nutrition. Am J Clin Nutr 2008;88: 1353–63. 51. Leitzmann MF, Stampfer MJ, Michaud DS, Augustsson K, Colditz GC, Willett WC, Giovannucci EL. Dietary intake of n23 and n26 fatty acids and the risk of prostate cancer. Am J Clin Nutr 2004;80:204–16. 52. De Ste´fani E, Deneo-Pellegrini H, Boffetta P, Ronco A, Mendilaharsu M. a-linolenic acid and risk of prostate cancer: a case-control study in Uruguay. Cancer Epidemiol Biomarkers Prev 2000;9:335–8. 53. Thie´baut AC, Chaje`s V, Gerber M, Boutron-Ruault MC, Joulin V, Lenoir G, Berrino F, Riboli E, Be´nichou J, Clavel-Chapelon F. Dietary intakes of x-6 and x-3 polyunsaturated fatty acids and the risk of breast cancer. Int J Cancer 2009;124:924–31. 54. Eussen SJ, Vollset SE, Hustad S, Midttun Ø, Meyer K, Fredriksen A, Ueland PM, Jenab M, Slimani N, Ferrari P, et al. Vitamins B2 and B6 and genetic polymorphisms related to one-carbon metabolism as risk factors for gastric adenocarcinoma in the European Prospective Investigation into Cancer and Nutrition. Cancer Epidemiol Biomarkers Prev 2010;19:28–38.