AJCN EB Program 2010 Early Registration
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Clin Nutr 89: 331-346, 2009. First published December 3, 2008; doi:10.3945/ajcn.2008.26834
 American Journal of Clinical Nutrition, doi:10.3945/ajcn.2008.26834
Vol. 89, No. 1, 331-346, January 2009
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
89/1/331    most recent
ajcn.2008.26834v1
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Saadatian-Elahi, M.
Right arrow Articles by Riboli, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Saadatian-Elahi, M.
Right arrow Articles by Riboli, E.
Agricola
Right arrow Articles by Saadatian-Elahi, M.
Right arrow Articles by Riboli, E.
© 2009 American Society for Clinical Nutrition

ORIGINAL RESEARCH COMMUNICATION

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 Nutrition1,2,3

Mitra Saadatian-Elahi, Nadia Slimani, Véronique Chajès, Mazda Jenab, Joëlle Goudable, Carine Biessy, Pietro Ferrari, Graham Byrnes, Philippe Autier, Petra HM Peeters, Marga Ocké, Bas Bueno de Mesquita, Ingegerd Johansson, Göran Hallmans, Jonas Manjer, Elisabet Wirfält, Carlos A González, Carmen Navarro, Carmen Martinez, Pilar Amiano, Laudina Rodriguez Suárez, Eva Ardanaz, Anne Tjønneland, Jytte Halkjaer, Kim Overvad, Marianne Uhre Jakobsen, Franco Berrino, Valeria Pala, Domenico Palli, Rosario Tumino, Paolo Vineis, Maria Santucci de Magistris, Elisabeth A Spencer, Francesca L Crowe, Sheila Bingham, Kay-Tee Khaw, Jakob Linseisen, Sabine Rohrmann, Heiner Boeing, Ute Noethlings, Karina Standahl Olsen, Guri Skeie, Eiliv Lund, Antonia Trichopoulou, Erifili Oustoglou, Françoise Clavel-Chapelon and Elio Riboli

1 From the International Agency for Research on Cancer, Lyon, France (MS-E, NS, VC, MJ, CB, PF, GB, and PA); the Institut Gustave Roussy, Villejuif, France (VC); the Université Lyon1, INSERM U870, Lyon, France (JG); the Julius Center for Health and Primary Care, University of Utrecht, Netherlands (PHMP); the Department of Epidemiology and Public Health, Faculty of Medicine, Imperial College London, London, United Kingdom (PHMP and ER); the National Institute of Public Health and the Environment, Bilthoven, Netherlands (MO and BBdM); the Department of Public Health and Clinical Medicine, Nutrition Research, Umeå, Sweden (IJ and GH); the Department of Surgery, Malmö University Hospital, Malmö, Sweden (JM); the Department of Clinical Sciences, Nutrition Epidemiology, Lund University, Malmö, Sweden (EW); the Unit of Nutrition, Environment and Cancer, Catalan Institute of Oncology, Barcelona, Spain (CAG); the Epidemiology Department, Murcia Health Authority and CIBER Epidemiologia y Salud Pùblica, San Sebastian, Spain (CN); the Andalusian School of Public Health and CIBER Epidemiologia y Salud Pùblica, Granada, Spain (CM); the Public Health Division of Gipuzkoa, Department of Health of the Basque Government, San Sebastian, Spain (PA); the Jefa de Sección de Información Sanitaria, Servicio de Salud Poblacional, Dirección General de Salud Pública, Consejeria de Salud y Servicios Sanitarios, Principado de Asturias, Spain (LRS); the Public Health Institute of Navarra and CIBER en Epidemiologia y Salud Pùblica, Pamplona, Spain (EA); The Danish Cancer Society, Institute of Cancer Epidemiology, Copenhagen, Denmark (AT and JH); the Department of Clinical Epidemiology, Aahrus University Hospital, Aalborg, Denmark (KO and MUJ); the Nutritional Epidemiology Unit, National Cancer Institute, Milan, Italy (FB and VP); the Dipartimento di Medicina Clinica e Sperimentale, Federico II University, Naples, Italy (MSdM); the Molecular and Nutritional Epidemiology Unit, CSPO Scientific Institute of Tuscany, Florence, Italy (DP); the Registro Tumori Ragusa, Ragusa, Italy (RT); the Servizio di Epidemiologia dei Tumori, Università di Torino and CPO-Piemonte, Turin, Italy (PV); the Cancer Epidemiology Unit, University of Oxford, Oxford, United Kingdom (EAS and FLC); the MRC Dunn Human Nutrition Unit and Centre for Nutritional Epidemiology in Cancer Prevention and Survival, Department of Public Health and Primary Care, University of Cambridge, United Kingdom (SB and K-TK); the German Cancer Research Centre, Heildelberg, Germany (JL and SR); the Department of Epidemiology, Postdam, Germany (HB and UN); the Institute of Community Medicine, University of Tromsø, Tromsø, Norway (KSO, GS, and EL); the Department of Hygiene and Epidemiology, University of Athens, Athens, Greece (AT and EO); and INSERM ERI-20, Institut Gustave Roussy, Villejuif, France (FC-C).

2 The work described herein was supported by the "Europe Against Cancer" Programme of the European Commission and other internal resources of the International Agency for Research on Cancer.

3 Reprints not available. Address correspondence to N Slimani, International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France. E-mail: slimani{at}iarc.fr.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Plasma phospholipid fatty acids have been correlated with food intakes in populations with homogeneous dietary patterns. However, few data are available on populations with heterogeneous dietary patterns.

Objective: The objective was to investigate whether plasma phospholipid fatty acids are suitable biomarkers of dietary intakes across populations involved in a large European multicenter study.

Design: A cross-sectional study design nested to the European Prospective Investigation into Cancer and Nutrition (EPIC) was conducted to determine plasma fatty acid profiles in >3000 subjects from 16 centers, who had also completed 24-h dietary recalls and dietary questionnaires. Plasma fatty acids were assessed by capillary gas chromatography. Ecological and individual correlations were calculated between fatty acids and select food groups.

Results: The most important determinant of plasma fatty acids was region, which suggests that the variations across regions are largely due to different food intakes. Strong ecological correlations were observed between fish intake and long-chain n–3 polyunsaturated fatty acids (r = 0.78, P < 0.01), olive oil and oleic acid (r = 0.73, P < 0.01), and margarine and elaidic acid (r = 0.76, P < 0.01). Individual correlations varied across the regions, particularly between olive oil and oleic acid and between alcohol and the saturation index, as an indicator of stearoyl CoA desaturase activity.

Conclusions: These findings indicate that specific plasma phospholipid fatty acids are suitable biomarkers of some food intakes in the EPIC Study. Moreover, these findings suggest complex interactions between alcohol intake and fatty acid metabolism, which warrants further attention in epidemiologic studies relating dietary fatty acids to alcohol-related cancers and other chronic diseases.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Assessment of dietary fat intake by commonly used dietary methods is in part limited by recall biases and inherent measurement errors (1). The latter are mostly due to factors such as the widespread distribution of different fats throughout the diet and the intrinsic difficulty that subjects have recognizing and reporting the consumption of hidden fats, which leads to underreporting, particularly by overweight and obese individuals (2, 3).

In light of these limitations, much research has been undertaken to identify biomarkers of habitual fat intake as alternative or complementary independent measures of dietary exposures to better understand the potential role of fats in the etiology of chronic diseases. However, no biomarker can accurately reflect the absolute habitual fat intake, largely because of endogenous fatty acid synthesis and complex fatty acid metabolism. Of the blood lipid fractions, triglycerides are influenced by the type and amount of fat consumed during recent meals; therefore, they are not the most appropriate markers of usual dietary fat intake. The fatty acid profile of serum, plasma, or erythrocyte membrane phospholipids reflects medium-term (weeks to months) intakes of some fatty acids and may be used as biomarkers of habitual intake of dietary fatty acids in large-scale epidemiologic studies (4, 5).

Although the plasma phospholipid composition of some fatty acids has been correlated with intake of various food groups in homogeneous populations (6–11), it is unclear whether such correlations may exist across different populations with heterogeneous dietary and lifestyle patterns. Thus, the main aim of this study was to analyze the relation between plasma phospholipid fatty acid concentrations and dietary intakes of major food groups within a cross-sectional substudy of the European Prospective Investigation into Cancer and Nutrition (EPIC)—a large cohort with >520,000 subjects from 23 different centers in 10 European countries (12, 13). Secondary objectives were to evaluate the potential of 2 different dietary assessment methods to predict plasma fatty acid concentrations at the population level and to examine the influence of different endogenous [age, sex, and body mass index (BMI; in kg/m2)] and exogenous (region and laboratory variables) factors on plasma fatty acid concentrations.


   SUBJECTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Subjects
The rationale and design of the EPIC Study was previously detailed (12). Briefly, the EPIC cohort consists of 500,000 subjects distributed among 23 centers in 10 European countries (Denmark, France, Greece, Germany, Italy, the Netherlands, Norway, Spain, Sweden, and the United Kingdom). Between 1992 and 1998, country-specific food-frequency questionnaires (FFQs; previous year's intake) and standardized lifestyle and personal history questionnaires were completed, and anthropometric data and blood samples were collected. In addition, detailed and standardized 24-h dietary recalls (24-HDRs) were collected from a stratified and representative subset (37,000 subjects) of the EPIC Study by using computerized EPIC-SOFT software (14, 15).

In each of the 23 recruitment centers, blood samples of ≥30 mL were drawn from most participants, stored at 5–10°C, protected from light, and transported to local laboratories for processing and preparing aliquots (16). The only exceptions were the EPIC-Oxford center (United Kingdom), where blood samples were collected from a network of general practitioners and transported to a central laboratory and centers in Sweden and Denmark and prepared as aliquots within 1 h of being collected. In all countries, except Denmark and Sweden, blood was separated into 0.5-mL fractions (serum, plasma, red blood cells, and buffy coat for DNA extraction). Each fraction was placed into straws, which were heat-sealed and stored under liquid nitrogen. In Denmark, 1.0-mL blood fraction aliquots were stored locally in Nunc tubes at –150°C under nitrogen vapor. In Sweden, samples were stored in freezers at –80°C.

For the purposes of the present analysis, 16 geographic areas (regions) were designated by grouping some of the 23 EPIC centers together: France (Paris and surrounding areas; in this study, most French participants came from northern France), Florence (central Italy), Varese/Turin (northern Italy), Ragusa/Naples (southern Italy), northern Spain (San Sebastian, Pamplona, and Oviedo), Granada (southern Spain), Murcia (southeastern Spain), Cambridge, the health conscious group in Oxford, the Netherlands (Utrecht and Bilthoven), Greece (Athens and other regions), Heidelberg (southwest Germany), Potsdam (former East Germany), Malmö (southern Sweden), Umeå (northern Sweden), and Denmark (Aarhus and Copenhagen).

Except for France, where only women were recruited, 100 men and 100 women were chosen from each of the 16 regions, which resulted in a total of 3100 subjects. In total, 3089 subjects were selected for participation in the study from those who completed the FFQs and 24-HDRs. Laboratory results for fatty acids were available for 3009 subjects, either because no biological material was available (4 aliquots missing) or because of the failure to extract fatty acids (76 subjects). Six additional subjects were dropped because they were not part of the EPIC central database, which left a total of 3003 subjects for the present study.

For Oxford, 65 of the 100 men selected were vegans (eating no animal products) and 35 were ovolactovegetarians (eating no meat or meat products); 88 of the 99 women selected were vegans and 11 were ovolactovegetarians. Subject selection was stratified by 4 age groups (45–49, 50–54, 55–59, and 60–64 y) with an equal number of men and women in each group. As far as possible, equal numbers of subjects were selected for each season in which the blood samples was collected.

Laboratory analysis
Straws containing citrated plasma were extracted from the EPIC biological bank, and samples of the same sex and age category were ordered randomly within analysis batches. Each batch included one subject from each participating center and one sample from a standard pool for the quality control (injected twice at the middle and end of each series). Samples from men and women were analyzed in separate batches by using a gas chromatography method that was detailed elsewhere (17). Briefly, on the day of analysis, straws of citrated plasma were unfrozen at room temperature, and total lipids were extracted from a 200-µL aliquot with 6 mL chloroform-methanol (2:1, vol:vol). Phospholipids were purified by adsorption chromatography on silica tubes and then transmethylated at room temperature to fatty acid methyl esters (FAMEs) with the use of 25 µL methyl-prep II. FAMEs were separated on a gas chromatograph (Hewlett-Packard, Palo Alto, CA) equipped with an on-column injector and a capillary column (length: 30 m; diameter: 0.32 mm; Supelco, Bellefonte, PA). Helium was used as a carrier gas at a flow rate of 1 mL/min, with nitrogen as a make-up gas for the flame ionization detector. Identification of individual FAMEs was based on the comparison of their relative retention time with authentic standard methyl ester fatty acids (Sigma, St Louis, MO). The relative amount of each fatty acid was expressed as the percentage of total area. Chromatograms were integrated automatically and then checked by a laboratory technician.

The analysis method allowed for the analysis of 22 individual fatty acids with a chain length between 14 and 22 carbons belonging to different fatty acid classes: saturated fatty acids (SFAs: 14:0, 15:0, 16:0, 17:0, 18:0, and 20:0), monounsaturated fatty acids (MUFAs: 16:1n–7, cis 18:1n–9, trans 18:1n–9, cis 18:1n–7, and 20:1n–9), n–3 polyunsaturated fatty acids (PUFAs: 18:3, 20:5, 22:5, and 22:6), and n–6 polyunsaturated fatty acids (PUFAs: 18:2, 18:3, 20:2, 20:3, 20:4, 22:4, and 22:5). Analytic quality control was carried out by the daily use of a standard quality-control plasma sample (n = 137). The CVs for the major fatty acids were as follows: 3.35% for 16:0, 1.91% for 18:0, 6.43% for cis 18:1n–9, 5.38% for trans 18:1n–9, 0.93% for 18:2n–6, 3.96% for 20:4n–6, 12.75% for 18:3n–3, 3.86% for 20:5n–3, and 4.55% for 22:6n–3. The within-day CVs, which were calculated by analyzing the fatty acid composition of 10 plasma samples during the same day, ranged from 0.29% for large peaks to 9.34% for the smallest peaks. All laboratory analyses were performed at the International Agency for Research on Cancer–World Health Organization.

Dietary data
Information on food consumption was assessed by FFQ for all subjects at study enrollment (18). Extensive quantitative or semiquantitative dietary questionnaires developed and validated locally were used in most centers, whereas a mixed method combining a semiquantitative FFQ and 14-d record was adopted in Malmö (19). The 24-HDR was obtained from a representative sample of each center by means of a standardized computerized interview program (EPIC-SOFT) (20). Individual portion sizes were estimated by using a common picture book throughout all centers (21) and other tools, such as depictions of household measurements, bread shapes, standard units, and portions. For this analysis, the same specific food groups to be correlated with biological measurements were considered for both the FFQ and the 24-HDR: meat (red and processed), fish (fatty, lean and total, including fish and shellfish), dairy products (milk, yogurt, and cheese), nuts and seeds, olive oil, margarine, and other vegetable oils.

Statistical analysis
All statistical analyses were performed using SAS statistical software (version 9.1; SAS Institute Inc, Cary, NC).

Plasma fatty acids
Means (±SD) plasma fatty acid concentrations by region and sex are presented on nontransformed data for descriptive purposes. Multivariate regression analysis was used to determine the influence of different factors (region, sex, BMI, age at blood collection, alcohol intake, season, smoking status, column, life time of column, and storage time since blood sampling) on plasma fatty acid concentrations; a backward elimination procedure was used to estimate the sum of squares for each independent variable, given other dependent variables in the model. The partial r2 value was calculated as the ratio of each variable's sum of squares, over the residual sum of squares of the model excluding that independent variable. This indicates the proportion of the variation explained by the model that can be attributed to each independent variable. The R2 value for the model was calculated by dividing the uncorrected total sum of squares by the corrected total sum of squares of the model.

Dietary data
The same food classification was used to classify foods reported through the FFQ and 24-HDR, after their quantities were expressed as finally consumed (ie, cooked and/or without wastes, where relevant) (20). For dietary questionnaires, the conversion of food quantities into nutrient intakes was obtained by means of country-specific nutrient databases. For the 24-HDR, used as reference calibration measurements in EPIC, an ad hoc nutrient database was compiled and standardized between the EPIC participating countries (22).

Ecological correlations
The crude means of select plasma phospholipid fatty acids and the crude mean intake of select food groups, as recorded by 24-HDR and FFQ, were calculated for each region and Spearman's rank-correlation coefficients were computed. The 24-HDR means were weighted by the season and day of the week (weekday compared with weekend) during which the recall took place to correct for day-to-day and seasonal variability of a single 24-HDR (23, 24). For the present analysis, 8 relevant food groups (dairy products, fish, meat, processed meat, nuts and seeds, olive oil, margarine, and other vegetable oils) were chosen on the basis of their high contribution to dietary fat, along with alcohol for its effect on fat metabolism. FFQ values for olive oil were unavailable for the EPIC-Umeå center, because its consumption was marginal: average intakes of 0.4 g in women and 1.2 g in men (25).

Individual correlations
Although a single 24-HDR is not appropriate for an analysis of correlations at the individual level, because of the large day-to-day variability in dietary consumption, individual Pearson correlation coefficients were calculated between plasma phospholipid fatty acid concentrations and the select food groups and for both dietary measurements (ie, 24-HDRs and FFQs) to confirm or not confirm observations made at the ecological level. All measurements were adjusted for sex, age, center, BMI, storage time, and total energy and weighted for season and day of the 24-HDR.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Plasma phospholipid fatty acid concentrations
Crude mean (±SD) plasma phospholipid fatty acid concentrations for men and women are provided in detail for all regions in Table 1. Five fatty acids accounted for more than two-thirds of total plasma phospholipid fatty acids: palmitic acid (16:0), linoleic acid (18:2n–6), stearic acid (18:0), arachidonic acid (20:4n–6), and oleic acid (cis 18:1n–9). Minor fatty acids were also measured, such as n–3 long-chain PUFAs: eicosapentaenoic acid (EPA, 20:5n–3), docosahexaenoic acid (DHA, 22:6n–3), and trans elaidic acid (18:1n–9).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Crude plasma phospholipid fatty acids by region1

 
We also determined the saturation index (SI) as the ratio of 18:0 to cis 18:1n–9 and the ratio of 16:0 to 16:1n–7. The SI is an indicator of activity of the rate-limiting enzyme stearoyl CoA desaturase ({Delta}9 desaturation), which transforms the SFAs 16:0 and 18:0 into the MUFAs 16:1n–7 (SIn–7) and cis 18:1n–9 (SIn–9), respectively. Overall, we observed weak differences between men and women: SFA and MUFA concentrations were higher, whereas n–6 and n–3 PUFA concentrations were lower in men than in women (Table 1 and Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Multiple regression analysis of predictors of main plasma phospholipid fatty acids in the cross-sectional European Prospective Investigation into Cancer and Nutrition (EPIC) Study1

 
Multiple regression analysis of predictors of main plasma fatty acids for men and women combined in all regions together showed that sex, BMI, season of blood drawing, age, smoking, alcohol intake, and storage time after blood sampling had small effects, even significant, compared with diet (Table 2). The more pronounced effect of alcohol intake concerned SFAs (16:0, 17:0, and 18:0), MUFAs (16:1n–7 and cis 18:1n–9), and the ratios of 16:0 to 16:1n–7 and of 18:0 to cis 18:1n–9. Geographic region appeared to be the most important factor of variability for fatty acids and had a strong effect on trans 18:1n–9, cis 18:1n–9, 18:3n–3, 20:5n–3, and 22:6n–3), but less of an effect on 18:0, cis 18:1n–7, and 18:3n–6 (Figure 1). The values represent the means of fatty acids, by center, expressed as a percentage of the total area for the fatty acids in each family, adjusted for age at blood donation, BMI, storage time, season of blood collection, smoking status, alcohol intake and sex after stratification by sex. SFAs did not vary across regions (Figure 1A). In contrast, MUFAs varied widely across regions with a clear geographic gradient, particularly for trans 18:1n–9 (Figure 1B). In general, southern Europe had the highest percentages of cis 18:1n–9 compared with northern Europe, except in Umeå where the percentage was almost the same as in Northern Spain. In contrast, the northern and central centers, particularly the 2 British centers, had the highest percentages of trans 18:1n–9 compared with the southern centers. Similarly to trans 18:1n–9, percentages of the essential fatty acid 18:3n–3 were globally higher in northern than in southern Europe, with the highest concentration in Sweden (Figure 1C). In contrast, the essential fatty 18:2n–6 did not vary substantially across regions (Figure 1C). The percentages of the n–3 long-chain PUFAs 20:5n–3 and 22:6n–3 were highest in Denmark and Sweden, lowest in the health conscious group in Oxford, and remained relatively constant across the other regions (Figure 1D). Concentrations of the n–6 long-chain PUFA 20:4n–6 did not vary widely across regions (Figure 1E).


Figure 1
View larger version (35K):
[in this window]
[in a new window]

  		FIGURE 1. Mean deviation (%) in the center-adjusted (by age at blood donation, storage time, BMI, season of blood collection, smoking status, alcohol intake, and sex) major fatty acids in each family from the mean of all regions. The reference circle of the radius (100%) indicated in each panel (A–E) corresponds to the global mean of all centers, and the spikes indicate the deviation of a given center mean from the mean of all centers. EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid.

 
Mean intakes of select food groups
Mean intakes of select food groups containing fatty acids, estimated from the 24-HDRs and FFQs, are presented in Table 3. Southern countries (Greece, Italy, and Spain) were characterized by a higher consumption of olive oil and a lower consumption of margarine compared with the other EPIC countries (France, Germany, Denmark, the Netherlands, the United Kingdom, and Sweden). Spain reported the highest fish and shellfish consumption, {approx}3–5 times higher than the lowest intake reported in the Netherlands and in Germany. Red and processed meat intake was the highest in the northern countries (Denmark, Sweden, Germany, the Netherlands, and France) and in northern Spain and the lowest in Greece. Dietary intake of dairy products was the highest in the Netherlands and the lowest in Italy. The highest consumption of alcohol was reported in Denmark, Germany, northern Spain, and Italy (Varese, Turin, and Florence) and the lowest in Athens and Umeå.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Intake of main food groups supplying fatty acids, estimated from 24-h dietary recalls (24-HDRs) and food-frequency questionnaires1

 
Correlations between plasma phospholipid fatty acid concentrations and select food groups at the ecological level
Correlations between mean plasma phospholipid fatty acid concentrations and mean dietary food groups are presented in detail, by center, in Table 4. Some strong correlations were found between individual fatty acids and intake of specific food groups, such as fish intake and 22:6n–3 concentration (Figure 2A), olive oil intake and cis 18:1n–9 concentration (Figure 2B), and margarine intake and trans 18:1n–9 concentration (Figure 2C).


View this table:
[in this window]
[in a new window]

  		TABLE 4. Ecological Spearman correlation coefficients between mean 24-h dietary recalls (24-HDRs) and food-frequency questionnaires (FFQs) for select food group intakes and mean plasma fatty acid concentrations by region (n = 16)1

 

Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 2. A–C: Ecological correlations between mean individual plasma fatty acid concentrations and mean intakes of specific food groups estimated from food-frequency questionnaires (FFQs) and 24-h dietary recalls (24-HDRs). Cam, Cambridge (United Kingdom); Den, Denmark; Flo, Florence (Italy); Fra, France; Gra, Granada (Spain); Gre, Greece (Athens); Hei, Heidelberg (Germany); Oxf, Oxford (United Kingdom); Mal, Malmö (Sweden); Mur, Murcia (Spain); Net, Netherlands; NSp, northern Spain; Pot, Potsdam (Germany); R/N, Ragusa/Naples (Italy); Ume, Umeä (Sweden); and V/T, Varese, (Milan)/Turin (Italy).

 
Some potential fatty acid patterns that appear to be associated with specific dietary groups are highlighted in Figure 3. For example, a high consumption of fatty fish is characterized by a high concentration of 22:6n–3, 20:5n–3, and 17:0 and by a low concentration of n–6 PUFAs in plasma phospholipids (Figure 3A). The same profile was associated with lean fish (Figure 3B). A high consumption of olive oil was characterized by a high concentration of cis 18:1n–9 and 17:0 and a low concentration of trans 18:1n–9 and 18:3n–3 in plasma phospholipids (Figure 3C). Other vegetable oils were associated with a high concentration of 17:0 and 20:4n–6 and with a low concentration of 18:3n–3 (Figure 3D). A high consumption of margarine was associated with a high concentration of trans 18:1n–9 and 18:3n–3 and with a low concentration of 20:4n–6 and 17:0 (Figure 3E). Dairy products were associated with a high concentration of 20:5n–3 and with a low concentration of 20:4n–6 (Figure 3F). No strong association was observed between meat intake and fatty acid concentrations (Figure 3G). Processed meat intake was associated with a high concentration of the SFAs 16:0 and 17:0 (Figure 3H). Intake of nuts and seeds was associated with a high concentration of 18:0 and 18:2n–6 and with a low concentration of trans 18:1n–9 (Figure 3I).


Figure 3
Figure 3
View larger version (88K):
[in this window]
[in a new window]

  		FIGURE 3. A–I: Ecological Spearman correlations between mean plasma phospholipid fatty acid concentrations and intakes of specific food groups. FFQ, food-frequency questionnaire; 24 hDR, 24-h dietary recall; t, trans; c, cis. (Continued on next page.)

 
We also determined correlations between plasma phospholipid fatty acid concentrations and their respective estimated dietary intakes available for all countries in the EPIC nutrient databases. Mean plasma phospholipid SFA concentrations, by center, were significantly correlated (although weakly) with estimated mean dietary intakes of SFAs (FFQ: r = 0.09, P < 0.0001; 24-HDR; r = 0.12, P < 0.0001), MUFA concentrations with estimated dietary intakes of MUFAs (FFQ: r = 0.25, P < 0.0001; 24-HDR: r = 0.18, P < 0.0001), and PUFA concentrations with estimated dietary intakes of PUFAs (FFQ: r = 0.23, P < 0.0001; 24-HDR: r = 0.18, P < 0.0001).

Correlations between plasma phospholipid fatty acid concentrations and select food groups at the individual level
Although it was not the initial purpose of this article, because of the use of a single 24-HDR for individual estimates, correlations be-tween food groups and plasma phospholipid concentrations performed at the individual level to confirm observations made at the aggregated level. The strongest correlations that we found at the ecological level remained at the individual level, but at much lower values: fatty fish intake and 22:6n–3 (FFQ: r = 0.29, P < 0.01; 24-HDR: r = 0.11, P < 0.01), olive oil and cis 18:1n–9 (FFQ: r = 0.26, P < 0.01; 24-HDR: r = 0.17, P < 0.01), and margarine and trans 18:1n–9 (FFQ: r = 0.08, P < 0.05; 24-HDR: r = 0.02, P < 0.05).

Some associations that were not found at the ecological level appeared at the individual level. Particularly, dairy product intake was associated with the sum of 15:0 + 17:0 (FFQ: r = 0.27, P < 0.01; 24-HDR: r = 0.21, P < 0.01) and with trans 18:1n–9 (FFQ: r = 0.17, P < 0.01; 24-HDR: r = 0.13, P < 0.01). Alcohol consumption was associated with 16:0 (FFQ: r = 0.30, P < 0.01; 24-HDR: r = 0.23, P < 0.01), cis 18:1n–9 (FFQ: r = 0.10, P < 0.01; 24-HDR: r = 0.09, P < 0.01), and 20:5n–3 (FFQ: r = 0.17, P < 0.01; 24-HDR: r = 0.14, P < 0.01), but was inversely associated with 18:0 (FFQ: r = –0.24, P < 0.01; 24-HDR: r = –0.18, P < 0.01) and with the ratio of 16:0 to 16:1n–7 (FFQ: r = –0.19, P < 0.01; 24-HDR: r = –0.18, P < 0.01) and of 18:0 to cis 18:1n–9 (FFQ: r = –0.17, P < 0.01; 24-HDR: r = –0.14, P < 0.01).

Interestingly, some correlations between diet and fatty acids varied by center. Olive oil intake was associated with cis 18:1n–9 concentration only in centers where olive oil consumption was high (Athens, Italy, and Spain), whereas no significant correlation remained in centers where olive oil consumption was low (Heidelberg, France, Potsdam, Denmark, the Netherlands, Malmö, Oxford, and Cambridge) (Figure 4). In centers where olive oil consumption was low, plasma phospholipid cis 18:1n–9 was not significantly correlated with other food groups, except for a positive correlation with meat intake in Cambridge. However, alcohol intake was associated with cis 18:1n–9 concentration only in centers where olive oil was low (Figure 4). Similarly, alcohol intake was inversely associated with the ratio of 18:0 to cis 18:1n–9 in countries where olive oil intake was low, whereas no association was found in countries where olive oil intake was high (data not shown).


Figure 4
View larger version (26K):
[in this window]
[in a new window]

  		FIGURE 4. A, B: Individual Pearson correlations between plasma phospholipid cis 18:1n–9 concentrations and intakes of olive oil and alcohol estimated from food-frequency questionnaires (FFQs) and 24-h dietary recalls (24-HDRs) at all centers.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
This study was the first large cross-sectional study providing data on plasma phospholipid fatty acid concentrations and their ecological and individual correlations with intake of select foods across European EPIC populations. Except for SFAs, "region" was the determinant factor of variability in plasma fatty acid concentrations, which suggests that the observed variations were likely largely due to different food habits and lifestyle. Strong ecological correlations indicated that plasma fatty acid profiles could be used as suitable biomarkers of dietary patterns. Finally, specific correlations between alcohol intake and some fatty acids suggest interactions with fatty acid metabolism.

Essential fatty acids that cannot be endogenously synthesized, such as 18:2n–6 and 18:3n–3, are expected to show direct relations between plasma concentrations and dietary intakes. Ecological analyses showed that 18:2n–6 correlated with intake of nuts and seeds, whereas 18:3n–3 correlated with margarine intake. The plasma concentration of 18:3n–3 was significantly higher in northern than in southern Europe, probably because of the addition of rapeseed oil to margarine in Swedish. The inverse ecological correlation between plasma 18:3n–3 and consumption of fish and olive oil might be related to particular dietary habits and regimens (ie, use of margarine compared with use of olive oil). The inverse correlation between 18:3n–3 and the estimated intake of vegetable oils might be explained by the fact that vegetable oils are a source of 18:2n–6, which competes with 18:3n–3 for absorption (26).

EPA and DHA are found mainly in fatty fish and fish oil, and their endogenous synthesis from {alpha}-linolenic acid is limited in vivo (27). The strongest association was found between fatty fish and 22:6n–3, whereas weaker associations were found with 20:5n–3, as previously reported in the EPIC-United Kingdom cohort (11). Furthermore, we found an association between alcohol intake and 20:5n–3, but not with 22:6n–3, at the individual level. An association with both fatty acids would have been expected from the positive correlation that we found between fish and alcohol intake (data not shown). The absence of an association with 22:6n–3 may suggest that alcohol enhances 22:6n–3 peroxidation or the retroconversion of 22:6n–3 to 20:5n–3 (28).

In a comparison of UK vegetarians and vegans with the other EPIC populations, the most striking difference observed was for 18:2n–6, which was higher in the plasma of the vegans, whereas n–3 PUFAs (20:5n–3 and 22:6n–3) and SFAs were lower. Other authors have reported similar findings (29, 30). The ratio of n–6 to n–3 PUFAs was 2–3 times that of vegans in the rest of the EPIC populations (13.71 and 11.65 for male and female vegans, respectively, compared with 3.89–6.42 for the other EPIC centers). The low concentrations of 20:5n–3 and 22:6n–3 in vegetarians and vegans may be further explained by the reduced conversion of 18:3n–3 to long-chain n–3 PUFAs because of the high ratio of n–6 to n–3 fatty acids. These findings agree with those reported earlier (29, 31). However, unlike Phinney et al (32), we did not observe a reduced concentration of 20:4n–6 in plasma associated with a vegetarian diet. In contrast with essential fatty acids, we found weaker associations between diet, SFAs, and MUFAs, probably because they are endogenously synthesized (33).

Of the saturates, fatty acids containing an odd number of carbon atoms (eg, 15:0 and 17:0) cannot be synthesized de novo and can only be derived from ruminant products (34). Many studies have reported a positive association between the 15:0 + 17:0 concentration in serum or plasma phospholipids and dairy products intake (9, 10, 35). In our study, the concentration of 15:0 + 17:0 was not associated with dairy products intake at the ecological level but an association was found at the individual level, which may be the consequence of a relatively low variability in the mean dietary intake of dairy products across centers. Our analyses at the individual level indicated that plasma phospholipid 15:0 + 17:0 may be considered a biomarker of dairy product intake.

Ecological analyses showed a strong correlation between the plasma concentration of cis 18:1n–9 and the estimated intake of olive oil. The highest concentration of cis 18:1n–9 was found in countries where the intake of olive oil, considered the major source of this fatty acid, was the highest (Greece, Italy, and Spain). The lowest concentration of cis 18:1n–9 along with a low intake of olive oil were found in France, probably because the majority of French participants originated from the northern part of the country, where olive oil consumption is low (25). However, we noted that the concentration of cis 18:1n–9 was high in Umeå, where the consumption of olive oil was low. Individual analyses showed that olive oil intake was strongly associated with plasma cis 18:1n–9 in centers where olive oil was high, >20 g/d (southern centers), whereas no significant association remained in centers where olive oil is low, <2 g/d (northern centers). In these centers, cis 18:1n–9 was not significantly associated with other food groups, which suggested that dietary contributors of cis 18:1n–9 may not be a determinant compared with endogenous hepatic synthesis from 18:0. In contrast, a high dietary intake of olive oil in southern Europe may be considered a major determinant of plasma cis 18:1n–9.

Individual analyses also showed that alcohol intake was positively associated with 16:0. The direct association between plasma phospholipid 16:0 and alcohol intake was also reported for erythrocyte membrane 16:0 in the Italian EPIC Study (10). As an underlying mechanism, alcohol may increase the activity of acetyl-CoA carboxylase and fatty acid synthase—the key enzymes in 16:0 synthesis (10). Alcohol was also inversely associated with the ratio of 16:0 to 16:1n–7 and with the ratio of 18:0 to cis 18:1n–9, as an indicator of activity of the stearoyl CoA desaturase, but these associations varied across centers. Inverse associations with alcohol were only found in centers where olive oil intake was low, whereas no significant inverse association remained in southern countries, where olive oil intake was high (data not shown). The inverse association with alcohol may reflect complex interactions with MUFA synthesis via {Delta}9 desaturation when the dietary supply of cis 18:1n–9 is low. As a potential mechanism, alcohol may enhance the conversion of SFAs to MUFAs via stearoyl CoA desaturase. When the dietary supply of cis 18:1n–9 is high, stearoyl CoA desaturase activity may be down-regulated via a feedback regulation. This finding warrants further interest because both alcohol consumption (36) and a low SI in serum (17, 37) or in erythrocyte membranes (38) have been associated with an increased risk of breast cancer.

We found a positive correlation between plasma trans 18:1n–9 concentration and margarine intake. trans Fatty acids are unsaturated fatty acids with at least one double bond in the trans configuration. Because humans do not synthesize trans fatty acids, the occurrence of these fatty acid isomers in blood depends on their availability in the diet. trans Fatty acids occur naturally in fat from ruminant-animal meat (mainly vaccenic acid, trans 18:1n–7), milk and dairy fat and unnaturally in industrially hardened vegetable oils (mainly elaidic acid, trans 18:1n–9) (39, 40). Dietary exposure to partially hydrogenated vegetable oils occurs through consumption of margarine and of industrially processed foods (41). A higher plasma concentration of trans 18:1n–9 was found in northern than in southern Europe, probably because of the higher consumption of margarine in northern Europe, which is the primary fat ingredient in a myriad of industrial products. Indeed, results of the EPIC dietary measurement showed that consumption of margarine ranges between 15 and 36.6 g/d in northern countries and between 0.4 and 4.5 g/d in southern countries of Europe (25). Thus, our data showed that the plasma phospholipid trans 18:1n–9 concentration can be considered a valid biomarker of margarine intake. The same data will be used in the near future to investigate the contribution of processed foods to the trans 18:1n–9 concentration.

Some fatty acids, particularly fatty acids not synthesized endogenously, are good biomarkers of diet estimated by FFQs and 24-HDRs. With few exceptions, correlations at the ecological level between plasma phospholipid fatty acids and diet are generally higher when estimated from 24-HDRs than when estimated from FFQs. This can be explained by the fact that, in contrast with FFQs, 24-HDRs are highly standardized across countries and provide a much higher level of detail and precision in estimating food intakes, particularly (added) fats. This finding is in line with findings reported in previous similar studies that used other independent biomarkers, such as urinary nitrogen and several plasma carotenoids (42, 43).

In summary, this study provides unique data on plasma phospholipid fatty acid concentrations and their correlation with food intakes in a large population with heterogeneous dietary habits. The dietary heterogeneity observed in European regions involved in this cross-sectional study will be a good basis in the future for investigating the relation between dietary fat intake and the incidence of cancer and cardiovascular diseases.


   ACKNOWLEDGMENTS
 
We thank all of the study participants for their cooperation and all of the persons who participated in the fieldwork studies in each EPIC center. We particularly thank Béatrice Vozar and David Achaintre for technical assistance with the laboratory assays.

The authors’ responsibilities were as follows—ER: coordinated the EPIC study in Greece (Greek Ministry of Health and Social Solidarity and Hellenic Health Foundation). MS-E, NS, VC, MJ, CB, PF, and GB: constituted the writing group, conducted the statistical analyses, and prepared the manuscript; VC: set up the laboratory method of analysis of plasma fatty acids and acted as an expert on fatty acid metabolism; NS: coordinated the development of the 24-HDR computerized program and the compilation of the EPIC Nutrient DataBase in collaboration with the EPIC centers and other external experts and supervised the preparation of the manuscript; all other authors contributed to and/or supervised the collection and analysis of dietary data and the collection of blood samples in the participating centers and/or provided comments and suggestions on the intermediate and final manuscripts. None of the authors had a personal or financial conflict of interest.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Bingham, SA, Luben, R, Welch, A, et al.. Are imprecise methods obscuring a relation between fat and breast cancer? Lancet 2003;362:212–4..[CrossRef][Medline]
  2. Johansson, L, Solvoll, K, Bjørneboe, GE & Drevon, CA. Under- and overreporting of energy intake related to weight status and lifestyle in a nationwide sample. Am J Clin Nutr 1998;68:266–74..[Abstract]
  3. Goris, AH, Westerterp-Plantenga, MS & Westerterp, KR. Undereating and underrecording of habitual food intake in obese men: selective underreporting of fat intake. Am J Clin Nutr 2000;71:130–4..[Abstract/Free Full Text]
  4. Riboli, E, Ronnholm, H & Saracci, R. Biological markers of diet. Cancer Surv 1987;6:685–718..[Medline]
  5. Kaaks, R, Riboli, E & Sinha, R. Biochemical markers of dietary intake. In: Toniolo, P, Boffetta, P, Shuker, DEG, Rothman, N, Hulka, B & Pearce, N, eds. Applications of biomarkers in cancer epidemiology. Lyon, France: International Agency for Research on Cancer, 1997:103–26..
  6. Dougherty, RM, Galli, C, Ferro-Luzzi, A & Iacono, JM. Lipid and phospholipid fatty acid composition of plasma, red blood cells, and platelets and how they are affected by dietary lipids: a study of normal subjects from Italy, Finland, and the USA. Am J Clin Nutr 1987;45:443–55..[Abstract/Free Full Text]
  7. Sasaki, S, Ushio, F, Amano, K, et al.. Serum biomarker-based validation of a self-administered diet history questionnaire for Japanese subjects. J Nutr Sci Vitaminol (Tokyo) 2000;46:285–96..[Medline]
  8. Kobayashi, M, Sasaki, S, Kawabata, T, Hasegawa, K, Akabane, M & Tsugane, S. Single measurement of serum phospholipid fatty acid as a biomarker of specific fatty acid intake in middle-aged Japanese men. Eur J Clin Nutr 2001;55:643–50..[CrossRef][Medline]
  9. Wolk, A, Furuheim, M & Vessby, B. Fatty acid composition of adipose tissue and serum lipids are valid biological markers of dairy fat intake in men. J Nutr 2001;131:828–33..[Abstract/Free Full Text]
  10. Fusconi, E, Pala, V, Riboli, E, et al.. Relationship between plasma fatty acid composition and diet over previous years in the Italian centers of the European Prospective Investigation into Cancer and Nutrition (EPIC). Tumori 2003;89:624–35..[Medline]
  11. Welch, AA, Bingham, SA, Ive, J, et al.. Dietary fish intake and plasma phospholipid n–3 polyunsaturated fatty acid concentrations in men and women in the European Prospective Investigation into Cancer-Norfolk United Kingdom cohort. Am J Clin Nutr 2006;84:1330–9..[Abstract/Free Full Text]
  12. 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):6S–14S..[Abstract/Free Full Text]
  13. Bingham, S & Riboli, E. Diet and cancer: the European Prospective Investigation into Cancer and Nutrition. Nat Rev Cancer 2004;4:206–15..[CrossRef][Medline]
  14. Slimani, N, Deharveng, G, Charrondiere, RU, et al.. Structure of the standardized computerized 24-h diet recall interview used as reference method in the 22 centers participating in the EPIC project. European Prospective Investigation into Cancer and Nutrition. Comput Methods Programs Biomed 1999;58:251–66..[CrossRef][Medline]
  15. Slimani, N, Kaaks, R, Ferrari, P, et al.. European Prospective Investigation into Cancer and Nutrition (EPIC) calibration study: rationale, design and population characteristics. Public Health Nutr 2002;5(6B):1125–45..[CrossRef][Medline]
  16. Riboli, E, Hunt, KJ, Slimani, N, et al.. European Prospective Investigation into Cancer and Nutrition (EPIC): study populations and data collection. Public Health Nutr 2002;5(6B):1113–24..[CrossRef][Medline]
  17. Chajès, V, Hultén, K, Van Kappel, AL, et al.. 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..[CrossRef][Medline]
  18. Kaaks, R & Riboli, E. Validation and calibration of dietary intake measurements in the EPIC project: methodological considerations. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol 1997;26(suppl 1):15S–25S..[Abstract/Free Full Text]
  19. Riboli, E, Elmstahl, S, Saracci, R, Gullberg, B & Lindgärde, F. The Malmö Food Study: validity of two dietary assessment methods for measuring nutrient intakes. Int J Epidemiol 1997;26:161S–71S..[Abstract/Free Full Text]
  20. Slimani, N, Ferrari, P, Ocke, M, et al.. Standardization of the 24-hour diet recall calibration method used in the European prospective investigation into cancer and nutrition (EPIC): general concepts and preliminary results. Eur J Clin Nutr 2000;54:900–17..[CrossRef][Medline]
  21. Van Kappel, AL, Amoyel, J, Slimani, N, Vozar, B & Riboli, E. EPIC-SOFT picture book for estimation of food portion sizes. Lyon, France: International Agency for Research on Cancer, 1994..
  22. Slimani, N, Deharveng, G, Unwin, I, et al.. The EPIC nutrient database project (ENDB): a first attempt to standardize nutrient databases across the 10 European countries participating in the EPIC study. Eur J Clin Nutr 2007;61:1037–56..[CrossRef][Medline]
  23. Maisey, S, Loughridge, J, Southon, S & Fulcher, R. Variation in food group and nutrient intake with day of the week in an elderly population. Br J Nutr 1995;73:359–73..[CrossRef][Medline]
  24. Fahey, MT, Sasaki, S, Kobayashi, M, Akabane, M & Tsugane, S. Seasonal misclassification error and magnitude of true between-person variation in dietary nutrient intake: a random coefficients analysis and implications for the Japan Public Health Center (JPHC). Cohort Study. Public Health Nutr 2003;6:385–91..[Medline]
  25. Linseisen, J, Bergström, E, Gafá, L, et al.. Consumption of added fats and oils in the European Prospective Investigation into Cancer and Nutrition (EPIC) centres across 10 European countries as assessed by 24-hour dietary recalls. Public Health Nutr 2002;5:1227–42..[Medline]
  26. Punchard, NA, Green, AT, Mullins, JG & Thompson, RP. Analysis of the intestinal absorption of essential fatty acids in vivo in the rat. Prostaglandins Leukot Essent Fatty Acids 2000;62:27–33..[Medline]
  27. Nikkari, T, Luukkainen, P, Pietinen, P & Puska, P. Fatty acid composition of serum lipid fractions in relation to gender and quality of dietary fat. Ann Med 1995;27:491–8..[Medline]
  28. Pawlosky, RJ & Salem, N. Perspectives on alcohol consumption: liver polyunsaturated fatty acids and essential fatty acid metabolism. Alcohol 2004;34:27–33..[CrossRef][Medline]
  29. Sanders, TA & Roshanai, F. Platelet phospholipid fatty acid composition and function in vegans compared with age- and sex-matched omnivore controls. Eur J Clin Nutr 1992;46:823–31..[Medline]
  30. Agren, JJ, Tormala, ML, Nenonen, MT & Hanninen, OO. Fatty acid composition of erythrocyte, platelet, and serum lipids in strict vegans. Lipids 1995;30:365–9..[Medline]
  31. Reddy, S, Sanders, TA & Obeid, O. The influence of maternal vegetarian diet on essential fatty acid status of the newborn. Eur J Clin Nutr 1994;48:358–68..[Medline]
  32. Phinney, SD, Odin, RS, Johnson, SB & Holman, RT. Reduced arachidonate in serum phospholipids and cholesteryl esters associated with vegetarian diets in humans. Am J Clin Nutr 1990;51:385–92..[Abstract/Free Full Text]
  33. Nakamura, MT & Nara, TY. Structure, function, and dietary regulation of delta6, delta5, and delta9 desaturases. Annu Rev Nutr 2004;24:345–76..[CrossRef][Medline]
  34. Wu, Z & Palmquist, DL. Synthesis and biohydrogenation of fatty acids by ruminal microorganisms in vitro. J Dairy Sci 1991;74:3035–46..[Abstract/Free Full Text]
  35. Smedman, AE, Gustafsson, IB, Berglund, LG & Vessby, BO. Pentadecanoic acid in serum as a marker for intake of milk fat: relations between intake of milk fat and metabolic risk factors. Am J Clin Nutr 1999;69:22–9..[Abstract/Free Full Text]
  36. Key, J, Hodgson, S, Omar, RZ, et al.. Meta-analysis of studies of alcohol and breast cancer with consideration of the methodological issues. Cancer Causes Control 2006;17:759–70..[CrossRef][Medline]
  37. Chajès, V, Thiébaut, AC, Rotival, M, et al.. Association between trans-monounsaturated fatty acids and breast cancer risk in the E3N-EPIC study. Am J Epidemiol 2008;167:1312–20..[Abstract/Free Full Text]
  38. Pala, V, Krogh, V, Muti, P, et al.. Erythrocyte membrane fatty acids and subsequent breast cancer: a prospective Italian study. J Natl Cancer Inst 2001;93:1088–95..[Abstract/Free Full Text]
  39. Sommerfeld, M. Trans unsaturated fatty acids in natural products and processed foods. Prog Lipid Res 1983;22:221–33..[CrossRef][Medline]
  40. Wolff, RL, Combe, NA, Destaillats, F, et al.. Follow-up of the delta4 to delta16 trans-18:1 isomer profile and content in French processed foods containing partially hydrogenated vegetable oils during the period 1995–1999: analytical and nutritional implications. Lipids 2000;35:815–25..[CrossRef][Medline]
  41. Aro, A, van Amelsvoort, J, Becker, W, et al.. trans Fatty acids in dietary fats and oils from 14 European countries: the TRANSFAIR study. J Food Comp Anal 1998;11:137–49..
  42. Slimani, N, Bingham, S, Runswick, S, et al.. Group level validation of protein intakes estimated by 24-hour diet recall and dietary questionnaires against 24-hour urinary nitrogen in the European Prospective Investigation into Cancer and Nutrition (EPIC) calibration study. Cancer Epidemiol Biomarkers Prev 2003;12:784–95..[Abstract/Free Full Text]
  43. Al-Delaimy, WK, Slimani, N, Ferrari, P, et al.. Plasma carotenoids as biomarkers of intake of fruits and vegetables: ecological-level correlation in the European Prospective Investigation into Cancer and Nutrition (EPIC). Eur J Clin Nutr 2005;59:1397–408..[Medline]
Received for publication August 13, 2008. Accepted for publication October 10, 2008.





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
89/1/331    most recent
ajcn.2008.26834v1
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Saadatian-Elahi, M.
Right arrow Articles by Riboli, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Saadatian-Elahi, M.
Right arrow Articles by Riboli, E.
Agricola
Right arrow Articles by Saadatian-Elahi, M.
Right arrow Articles by Riboli, E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS