Biomarker-based evaluation of two 24-h recalls for comparing usual fish, fruit and vegetable intakes across European centers in the EFCOVAL Study

Article metrics

Abstract

Background/Objectives:

A standardized methodology is important to enable consistent monitoring of dietary intake across European countries. For this reason, we evaluated the comparability of the assessment of usual food intake collected with two non-consecutive computerized 24-h dietary recalls (24-HDRs) and a food propensity questionnaire (FPQ) among five European centers.

Subjects/Methods:

Two 24-HDRs using EPIC-Soft (the software developed to conduct 24-HDRs in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study) were performed to determine fish, fruit and vegetable (FV) consumed by 600 adults in Belgium (BE), the Czech Republic (CZ), France (FR), the Netherlands (NL) and Norway (NO) in a validation study. An FPQ was used to identify non-consumers. Information from the 24-HDRs and FPQ were used to estimate individual usual food intake by the Multiple Source Method (MSM). Blood samples were drawn to determine fatty acids in phospholipids and serum carotenoids as biomarkers of fish, and FV intake, respectively.

Results:

The pooled correlation between usual fish intake and eicosapentaenoic acid plus docosahexaenoic acid in phospholipids was 0.19 in men and 0.31 in women (P for heterogeneity >0.50) and center-specific correlations ranged between 0.08 (CZ) and 0.28 (BE and NO) in men, and between 0.19 (BE) and 0.55 (FR) in women. For usual FV intake, the pooled correlation with serum carotenoids was 0.31 in men and 0.40 in women (P for heterogeneity >0.10); the center-specific correlations varied between 0.07 (NO) and 0.52 (FR) in men, and between 0.25 (NL) and 0.45 (NO) in women.

Conclusion:

Two standardized 24-HDRs using EPIC-Soft and an FPQ appeared to be appropriate to rank individuals according to their fish and FV intake in a comparable way among five European centers.

Introduction

Dietary data from national food consumption surveys are useful to develop and evaluate policies on nutrition and food safety. In Europe, national food consumption data are important to assess the variability of food patterns among different countries. However, European countries performing national surveys use different methodologies such as 24-h dietary recalls (24-HDRs) and food diaries to collect dietary data (EFSA, 2009). In addition, differences exist in a number of aspects such as the food classification system used across countries. For instance, olives can be considered as a fruit in one food classification, and as a vegetable in another (Ireland et al., 2002).

European countries are expected to provide similar dietary indicators if harmonized food consumption data are collected in future national surveys (Brussaard et al., 2002a). For this reason, the European Food Consumption Survey Method (EFCOSUM) Consortium recommended the collection of food consumption data using two non-consecutive standardized 24-HDRs as the most appropriate method in future pan-European surveys (Brussaard et al., 2002b). Furthermore, the consortium recommended the use of EPIC-Soft (the software developed to conduct 24-HDRs in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study) for standardization, and defined a set of dietary components, including, besides specific nutrients, vegetables, fruits, bread, fish and shellfish (Biro et al., 2002; Slimani and Valsta, 2002; Steingrimsdottir et al., 2002; Brussaard et al., 2002b) to serve as nutritional indicators.

Because the use of a standardized and valid methodology is crucial to enable consistent monitoring of diet across European countries, the European Food Consumption Validation (EFCOVAL) Consortium aimed to further develop and validate the methodology proposed for pan-European dietary monitoring. To that end, our previous work showed that two non-consecutive days of dietary intake collected with 24-HDRs (EPIC-Soft) were considered sufficiently valid for comparing usual protein and potassium intake among five European centers (Crispim et al., 2011). In the present study, we intended to further evaluate the dietary intake collected with respect to the comparability of food group assessment across different European populations. A food propensity questionnaire (FPQ) was included in the assessment to offer covariate information in complementing the 24-HDRs during the estimation of usual intake of food groups (Subar et al., 2006).

Assessment of intake of fruits and vegetables (FVs) and fish and shellfish can be evaluated using, serum carotenoids (Yeum et al., 1996; Polsinelli et al., 1998; Macdonald et al., 2009) and n-3 fatty acids, respectively, in, for example, phospholipids (Leaf et al., 1995; Katan et al., 1997; Zock et al., 1997) as concentration biomarkers. Concentration biomarkers are related to dietary intake, but not as directly as recovery biomarkers because their concentrations are the result of complex metabolic processes (Freedman et al., 2010). Therefore, their use in validation studies is restricted to their associations, commonly as correlations, with self-reported dietary intakes. The strength of these correlations is often lower (<0.6) than that of recovery biomarkers (Jenab et al., 2009).

This paper aims to evaluate and compare the assessment of ranking of individuals according to their usual fish and FV consumption estimated with two non-consecutive standardized 24-HDRs and an FPQ between five selected centers in Europe, using fatty acids in phospholipids and serum carotenoids as biomarkers of intake, respectively.

Subjects and methods

Study population

The study population consisted of 297 men and 303 women, in the age group between 45 and 65 years, from five selected centers from Belgium, the Czech Republic, France (Southern part), the Netherlands and Norway. These centers were chosen to represent the large diversity of food patterns across Europe. Participants were recruited by convenience sampling, and were healthy individuals representing all educational levels. Eligible participants were able to read and speak the national language, not following prescribed dietary therapy, not pregnant or lactating and not enrolled in another study at the same period. In addition, we did not allow subjects in the study who were donating blood or plasma during or <4 weeks before the study, institutionalized persons or more than one member of the same household. More details about the study populations, including recruitment and sampling procedures are described elsewhere (Crispim et al., 2011).

Study design

The period of data collection was from April to July 2007 in the Netherlands and from October or November 2007 to April 2008 in the other four centers. Ethics committees in each center approved the research protocol, and participants signed an informed consent. At the beginning of the study, each participant filled out a screening and a general questionnaire, with questions about lifestyle and food habits, including type and frequency of supplements used during the previous 3 months. Participants were then weighed and had their height measured in the study centers following standardized procedures. They also underwent a non-fasting venipuncture. Thereafter, we collected two non-consecutive 24-HDRs with approximately 1 month gap in between. The time interval between blood sampling and the first 24-HDR was on average less than a week for all centers, except in the Czech Republic where the average was 2 weeks.

Dietary data

We collected the two 24-HDRs using EPIC-Soft, version 9.16 (Slimani et al., 1999, 2000). In brief, EPIC-Soft is a computer-assisted dietary tool that follows a standardized procedure to minimize measurement errors when describing, quantifying, probing and calculating food intakes across countries (Slimani et al., 1999). The two 24-HDRs were collected using two modes of administration: one through phone and one by conducting face-to-face interviews. A randomization schedule was created to consider a random order of the two modes of administration, as well as the inclusion of all days of the week equally among the subjects. This randomization allowed the same person to have the same day of the week recalled for both interviews by chance.

Interviewers in each center were nutritionists or dietitians who were trained by qualified local trainers in interviewing skills and working with EPIC-Soft. Centers were allowed to organize their data collection in the same way as they would in a future performance of their national monitoring survey. For example, dietary recalls in Belgium, the Czech Republic and the Netherlands were not conducted on Sundays and, therefore, intakes from Saturday were recalled 2 days later, that is, on Mondays. Furthermore, interviewees were permitted to check food packages and household measures in their home for detailed information during the phone interview, whereas this was not possible during the face-to-face interview at the study center. All centers used an existing version of EPIC-Soft, which had already been used in a national survey or within the EPIC study, except the Czech Republic for which a new version was developed. Methods of estimation of portion size included household measures, weight/volume, standard units and portions, as well as photographs in a picture book. Furthermore, dietary supplement-use information of the recalled day was collected at the end of the 24-HDR interview. If a supplement was taken, subjects reported on the physical state (for example, capsule), the number of units per consumption occasion and the frequency. If known, the brand name was also reported. In addition, an FPQ including one question per food group was used to identify frequency of usual consumption of fish, FVs over the past year.

Food groups were defined as suggested by European Food Consumption Survey Method (Ireland et al., 2002). To this end, the foods, as reported by the recalls, were regrouped by including or excluding specific subgroups of the EPIC-Soft food classification (Slimani et al., 1999). Fruit intake was defined not to include nuts, seeds, olives and fruit juices other than freshly squeezed juices. Vegetable intake was defined to include herbs but not pulses and potatoes. Fish intake was defined to include shellfish. Fish was classified in lean fish (<4 g of fat/100 g of edible part such as cod, tuna, tilapia and carp) or fatty fish (4 g of fat/100 g of edible part such as salmon, herring and mackerel) using country-specific food composition tables.

Venipuncture and biomarkers

We provided participants with guidelines to consume a low-fat breakfast before blood sampling. We requested subjects to rest before a trained lab technician drew blood (2 × 9 ml) from the antecubital vein. The blood was then allowed to clot for 30 min at room temperature (20–22 °C) and centrifuged for 15 min at 1200 g. Serum samples from each subject were aliquoted into cryotubes for storage at −80 °C, until shipment on dry ice to the central laboratory at Wageningen University, where analyses were carried out.

After thawing and mixing the samples, fatty acids in the phospholipid fraction were measured by extracting and separating the lipid classes. Briefly, the phospholipid fraction was separated from the other lipid classes on an aminopropyl column according to the procedure described by Kaluzny et al. (1985). Fatty acid methyl ester profile was prepared according to the procedure described by Metcalfe et al. (1966). Serum carotenoids were analyzed as described by Khan et al. (2007). This method does not adequately separate lutein and zeaxanthin; consequently, these two carotenoids are presented together. Total cholesterol was measured spectrophotometrically on a Synchron LX20 clinical analyzer (Beckman Coulter, Mijdrecht, The Netherlands).

The fatty-acid composition of phospholipids was used as concentration biomarker of fish intake, namely the percentage of eicosapentaenoic acid (EPA; C20:5n-3) and docosahexaenoic acid (DHA; C22:6n-3) in relation to the total area of measured fatty acids (36 fatty acids). The sum of serum carotenoids, including α-carotene, β-cryptoxanthin, β-carotene, lutein and zeaxanthin, was used as marker of FV intake. To further explore the correlations across centers, both α-carotene and β-carotene were used as biomarkers of FVs. Similarly, β-cryptoxanthin was used as a biomarker of fruit intake (Polsinelli et al., 1998; Jansen et al., 2004) and lutein plus zeaxanthin of vegetable intake (Jansen et al., 2004; Al-Delaimy et al., 2005b). The intraassay precision, expressed as coefficient of variation, of EPA and DHA was <4% and of individual carotenoids between 5 and 8%.

Supplement use

Supplement users were identified as those who reported taking any supplements containing EPA, DHA or carotenoids on one of the recalled days or during the past 3 months according to the general questionnaire. To identify the presence of fatty acids and carotenoids in the reported supplements, we (1) searched websites of companies, (2) visited drugstores and (3) searched other sources such as national databases.

Statistical analyses

The statistical analyses were carried out separately for men and women. For evaluating the ranking of individuals, we calculated Pearson's correlation coefficients between the average intake of foods groups, based on the 2 days and their respective biomarkers per center. In addition, adjusted Pearson's correlation coefficients were estimated between usual intake of food groups and respective biomarker using partial correlations. For adjusted correlations, we used usual intake corrected for within-person variability together with the information from the FPQ (Adjusted1), as estimated by the Multiple Source Method (MSM; German Institute of Human Nutrition, 2009). We further corrected for the following covariables that were expected to be associated with the intake or excretion on the basis of pre-existing knowledge: age, body mass index, educational level, alcoholic beverage intake and smoking status (Adjusted2). Fruit and vegetable intake analyses were also corrected for total serum cholesterol (Brady et al., 1996). For the calculation of the correlations, intake of foods and concentrations of biomarkers were log-transformed to improve normality of the observed distributions. Considering the fact that improvements in the normality of the distribution may have not been achieved, Spearman's correlations were also computed. However, only when conclusions based on Pearson's and Spearman's correlations differed, we presented the latter. Confidence intervals of the correlations were obtained using Fisher's Z-transformation (Kleinbaum et al., 2007).

The MSM is a statistical method to estimate usual dietary intake of nutrients and foods, including episodically consumed foods, for populations as well as individuals. In contrast to many other statistical methodologies, MSM first estimates individual usual intakes rather than constructing directly the population distributions of usual intake. The method can make use of covariate information such as consumption frequency information from an FPQ to improve the modeling of consumption probability and intake amount (German Institute of Human Nutrition, 2009).

Pooled correlations of the five centers were calculated by first converting correlations into a standard normal metric (Fisher's r-to-Z transformation). Next, the pooled average was calculated, in which each transformed correlation coefficient was weighted by its inverse variance, followed by the back transformation (Kleinbaum et al., 2007). Cochrane Q-test was used for testing heterogeneity of the pooled correlation (Field, 2005).

The estimated intake did not include the amounts of EPA, DHA or carotenoids, originating from supplement use. To help interpreting the main results, biomarker levels were presented separately for the total sample, users and non-users of supplements. Given the small number of subjects in each group, men and women were grouped together to optimize this part of the analysis.

Analyses were performed using SAS statistical package, version 9.1 (SAS Institute Inc., Cary, NC, USA).

Results

With regard to the characteristics of our study population, the mean body mass index of the French men and women was somewhat lower than those of the other four centers (Table 1). A larger prevalence of smokers was observed in Czech men (33%) and Norwegian women (23%) than in subjects of the other centers. Moreover, subjects with a low educational level were less represented than subjects with a moderate or high educational level, especially in Norwegian men. Belgian men reported the highest intake of alcoholic beverages (average of 30.2 g/day) and Czech women the lowest (average of 6.3 g/day). Furthermore, the total serum cholesterol concentration of the subjects did not vary substantially across the five centers in both genders.

Table 1 Characteristics of subjects from the five European centers in the EFCOVAL Study

In all centers, each day of the week was represented by between 12 and 17% of the 24-HDRs, except in France, where Saturday was less representative (8%) and Thursday more (19%), and in the Czech Republic, where almost 19% of the interviews were about the intake of a Sunday. For all centers, the interval between the first and the second 24-HDRs was at least 3 weeks.

Mean fish intake was highest in the Norwegian center; it was 3–4 times higher than the mean intakes in the Czech Republic or the Netherlands (Table 2). Similarly, the highest mean percentage of EPA plus DHA in phospholipids was seen in Norway and the lowest in the Czech Republic and the Netherlands. A low proportion of fatty to lean fish intake (excluding fish products) was observed in the Czech and French men (ratio 0.3) as compared with the other three centers, in which the ratio ranged between 0.9 and 1.1. A high proportion of fatty to lean fish (ratio 4.3) was observed in Dutch women. Shellfish and roe products contributed between 13 and 26% of total fish intake in Belgium and France, whereas the Czech Republic did not report any consumption of it (data not shown). The percentage of fish consumers identified by 24-HDRs was lower (Table 2) than that identified by FPQ, which showed nearly 95% of consumers in all centers (data not shown).

Table 2 Intakes of fish, fruits and vegetables estimated from the 2 × 24-HDRs and related biomarkers (mean±s.e.) of 600 participants in the EFCOVAL Study

The lowest crude correlation between fish intake and the biomarker (Table 3) was observed in the Czech Republic in both genders (r=−0.04 in men and 0.24 in women). When we analyzed usual fish intake by adjusting intakes for within-person variability and including FPQ data (see Table 3, Adjusted1), evident improvement of correlations was seen across the centers, with the exception of the correlation coefficients in Belgian and Czech women that decreased, respectively, from 0.34 to 0.19 and from 0.24 to 0.21. Further adjustment of the correlations for possible confounders did not explain the differences across centers (see Table 3, Adjusted2). Nevertheless, although the adjusted correlation for fish intake was still considerably lower in Czech men (r=0.08) than in the other centers, no statistically significant heterogeneity of correlations was found between the centers (P>0.20 in both gender).

Table 3 Pearson's correlation coefficients (r) with confidence intervals (CIs) between intakes of fish, fruit and vegetable from 24-HDRs and related biomarkers of participants in the EFCOVAL Study

The largest average intake of both FVs was seen in France for both men and women (Table 2). Although the lowest fruit intake was reported in Belgium for men and in Norway for women, the lowest vegetable intake was observed in the Czech Republic and Norway for both genders. In addition, there were large differences in the types of FV consumed across centers for both men and women. Cooked FVs were less consumed by Czech subjects (25%) and more by French (45%). The average amounts of citrus fruits consumed were larger in the Czech Republic and France than in the other three centers. Leafy vegetables were clearly less consumed in the Czech Republic and Norway than in Belgium, France and the Netherlands. The percentage of FV consumers identified by the 24-HDRs was the same as that identified by the FPQ, and was nearly 100%. In relation to the biomarker, the highest mean concentration of carotenoids was observed in France and the lowest in the Czech Republic.

Crude correlation coefficients of FV intake with the sum of carotenoids were between 0.05 in the Norwegian men and 0.54 in the Czech men (Table 3). A smaller range of correlations was observed in women (0.33 in the Czech Republic to 0.49 in Belgium). Pearson's correlations, based on usual intakes, as estimated by the MSM method, barely differed from the crude ones (Adjusted1). Adjusted correlations including possible confounders varied in different directions and did not explain the differences across the centers (Adjusted2). The adjusted correlation of FV intake in the Norwegian men was 0.07, whereas all other centers presented correlations ranging from 0.16 to 0.52. However, we did not identify deviating correlations when assessing the heterogeneity of the pooled correlations (P>0.10 for all comparisons). Overall, the correlations of the combined intake of FVs with the sum of carotenoids were higher than that of FVs separately, especially for fruit. Correlations between FV intake with α-carotene were higher than that with the sum of carotenoids in some sub-populations but lower in others such as in the Dutch men, who happened to have the lowest consumption of carrots (results not shown). When using β-carotene as a biomarker of FV intake, correlations were often lower than the sum of carotenoids, particularly in the Netherlands. The correlations between vegetable intake and luteine plus zeaxanthine, and the correlations between fruit intake and β-cryptoxanthin did not explain the low correlation observed in Norway. Nevertheless, when using Spearman's correlations, there was an improvement of those correlations in Norway. Furthermore, after including all types of juices in the FV group, the correlations modestly increased between FV intake and the sum of carotenoids in some centers, but not for all (data not shown). The major changes were seen in Norway for men, where for instance the correlation between fruit intake and β-cryptoxanthin increased from 0.03 to 0.32.

The percentage of fish oil supplement users was high in Norway (63%) as compared with the other four centers (<14%; Table 4). In line with this, the percentage of EPA plus DHA in phospholipids of subjects, who reported not taking any fish oil supplement, was substantially lower than that of the supplement users and the total group in Norway. Supplements containing carotenoids were less often consumed than those with fish oil, with the highest number of users in the Czech Republic and the Netherlands (11 subjects each) and the lowest in Norway (1 subject). As a result, mean serum carotenoid concentrations of non-supplement users were similar to those of the total group.

Table 4 Biomarker levels (mean±s.e.) of the total sample, users and non-users of specific supplements in the EFCOVAL Study

Discussion

We compared the assessment of usual fish and FV consumption of adults estimated with two non-consecutive standardized 24-HDRs in combination with a FPQ among five centers in Europe. Overall, we observed weak-to-moderate associations between fish and FV intake and respective biomarkers. In men, correlations for fish intake in the Czech Republic and for FV intake in Norway were distinctly lower than those in the other centers. In women, the correlations across centers were rather comparable.

One of the major strengths of this study is the replicate collection of 24-HDRs, allowing the application of statistical adjustment to obtain an estimate of the individual usual intake. Welch et al. (2006) have shown that one 24-HDR was less consistent in providing an association between fish intake and serum fatty acids than food frequency questionnaires or a 7-day diary. This can be explained by the fact that 24-HDRs are not able to reflect usual intakes of infrequently consumed foods. Indeed, we have observed that correlations substantially improved when considering usual intakes including FPQ data for fish. Another important strength of this study is the unique setting of data collection that has provided standardized dietary intake and biomarker information for different countries.

One of the limitations of our study is that given the limited number of participants, large confidence intervals were observed. The sample size may also have limited the interpretation of the Cochrane Q-test. We found no statistically significant heterogeneity between correlations, but this could be caused by the relatively small sample size. Nevertheless, this is not very likely because the observed P-values for that test were rather high, especially in the assessment of fish intake (P>0.50). In addition, data collection was performed in a different season in the Netherlands than in the other four centers. Considering the fact that carotenoid contents in FVs may differ between seasons (Maiani et al., 2009), this could have led to a different performance of the method in the Dutch population. Nevertheless, we expect that both intake and biomarker assessment have been affected, thus minimizing the possible influence of seasonality on the correlations. Another potential limitation may be that the five center populations are not representative of their respective country populations, because they can be expected to consist of health-conscious subjects. This hampers the extrapolation of our results to the general population. Furthermore, the individual usual intakes of foods estimated with MSM can be questioned. A study by Souverein et al. (2011) showed that when applying methods such as MSM to groups of small sample size, the estimates of usual intake distributions are highly uncertain. However, the accuracy of individual usual intakes estimated by MSM remains unclear, as it has not been investigated.

Apart from sampling variation, the differences in correlations between centers can be attributed to differences in the range of food intake, different compositions of the foods consumed and the presence of other determinants or modifiers of the concentration biomarkers. The range of food intakes differed across centers, especially for fish intake. The low correlation for fish in Czech men may be explained by their very low amounts of fish consumption (on average 20 g/day) by only few subjects (26%). In the Czech population, the fish consumption is traditionally very low and our finding agrees with the low intake of 13 g/day according to a household budget survey (Dofkova et al., 2001). In terms of differences in types of foods consumed, studies have shown that the correlation between fatty fish with serum and plasma fatty acids was stronger than that for total fish (Hjartaker et al., 1997; Welch et al., 2006). We were unable to present these correlations for our data given the high number of non-consumers for fish subgroups and the lack of specific FPQ data. Even so, we observed a very low ratio between fatty and lean fish intake in Czech men and in France. However, in France, a considerable amount of shellfish and roe products was reported, which also contributes to n-3 fatty acids. The low consumption of shellfish and roe products together with the low ratio between fatty and lean fish intake might explain why the Czech Republic presented a very low correlation between fish intake and the biomarker, and France did not. Furthermore, substantial differences were observed in the types of FVs consumed across centers. Because the contents and bioavailability of carotenoids in foods can differ depending on harvest conditions, degree of maturity, storage and physical state (Maiani et al., 2009), populations with different FV intakes, being more represented by one specific carotenoid than another, may have different carotenoid profiles as well. As a consequence, the sum of carotenoids may not sensibly represent the carotenoid content of FVs of a specific population. We did observe different correlations across centers when using specific carotenoids as biomarkers of specific FV intake, but these did not substantially explain any observed differences in the comparison of FV intake vs sum of carotenoids across centers. Furthermore, other dietary sources may contribute to n-3 fatty acids or carotenoids in the blood. For example, some oils are rich sources of α-linolenic acid, which may, to a low extent, be converted to EPA and DHA (Connor, 2001), and colored foods, such as cheeses, can contain carotenoids (Gordon and Bauernfeind, 1982; Jones et al., 2005). Non-fresh FV juices, which were not included in the FV group, and fortified foods may also contribute to concentrations of carotenoids. These sources may partly explain the low associations between intake and biomarker, as observed in Norway.

In addition, the interpretation of our results demands understanding of aspects influencing not only the assessment of intake, but also that of the biomarker. Concentration biomarkers do not reflect absolute intake and their quantitative relationship with diet may vary between populations, depending on the presence and relative impact of determinants, such as genetic variation and lifestyle factors (Kaaks et al., 1997; Jenab et al., 2009). For instance, smoking and alcohol consumption have been inversely associated with levels of n-3 fatty acids (Simon et al., 1996; Galan et al., 2005; Pawlosky et al., 2007) and serum carotenoids (Rock et al., 1999; Northrop-Clewes and Thurnham, 2007). Nevertheless, when adjusting our correlations for a number of potential confounders such as alcohol intake, smoking status and total serum cholesterol concentrations, the outcomes were quite similar.

In addition, the association between food intake and biomarkers may have been influenced by the use of supplements (Hjartaker et al., 1997; Welch et al., 2006). Although the Norwegian sample was not the center with the most deviating association between fish intake and its biomarker, the percentage EPA plus DHA in phospholipids in supplement users was markedly higher than that in non-supplement users and the total population in Norway. Nevertheless, it is questionable whether participants were able to recall the exact type and brand name of their supplements (Skeie et al., 2009). Therefore, a degree of uncertainty remains in the evaluation of supplement use in relation to food intake assessment and in the explanations of differences in correlations across centers.

The correlations in this paper are consistent with the results of other studies that have found weak to moderate correlations between fish intake and n-3 fatty acids in the blood (Welch et al., 2006; Saadatian-Elahi et al., 2009) and between FV and serum carotenoids (Resnicow et al., 2000; Al-Delaimy et al., 2005a) when using 24-HDRs (r for fish intake between 0.11 and 0.22 and for FV intake between 0.30 and 0.42).

Despite the limitations, we conclude that two standardized 24-HDRs using EPIC-Soft in combination with an FPQ appeared to be appropriate to rank subjects according to their usual fish and FV intake within the five European centers in a comparable manner.

References

  1. Al-Delaimy WK, Ferrari P, Slimani N, Pala V, Johansson I, Nilsson S et al. (2005a). Plasma carotenoids as biomarkers of intake of fruits and vegetables: individual-level correlations in the European Prospective Investigation into Cancer and Nutrition (EPIC). Eur J Clin Nutr 59, 1387–1396.

  2. Al-Delaimy WK, Slimani N, Ferrari P, Key T, Spencer E, Johansson I et al. (2005b). Plasma carotenoids as biomarkers of intake of fruits and vegetables: ecological-level correlations in the European Prospective Investigation into Cancer and Nutrition (EPIC). Eur J Clin Nutr 59, 1397–1408.

  3. Biro G, Hulshof KF, Ovesen L, Amorim Cruz JA (2002). Selection of methodology to assess food intake. Eur J Clin Nutr 56 (Suppl 2), S25–S32.

  4. Brady WE, Mares-Perlman JA, Bowen P, Stacewicz-Sapuntzakis M (1996). Human serum carotenoid concentrations are related to physiologic and lifestyle factors. J Nutr 126, 129–137.

  5. Brussaard JH, Johansson L, Kearney J (2002a). Rationale and methods of the EFCOSUM project. Eur J Clin Nutr 56 (Suppl 2), S4–S7.

  6. Brussaard JH, Lowik MR, Steingrimsdottir L, Moller A, Kearney J, De Henauw S et al. (2002b). A European food consumption survey method--conclusions and recommendations. Eur J Clin Nutr 56 (Suppl 2), S89–S94.

  7. Connor WE (2001). n-3 Fatty acids from fish and fish oil: panacea or nostrum? Am J Clin Nutr 74, 415–416.

  8. Crispim SP, de Vries JH, Geelen A, Souverein OW, Hulshof PJ, Lafay L et al. (2011). Two non-consecutive 24 h recalls using EPIC-Soft software are sufficiently valid for comparing protein and potassium intake between five European centers—results from the European Food Consumption Validation (EFCOVAL) study. Br J Nutr 105 : 447–458.

  9. Dofkova M, Kopriva V, Resova D, Rehurkova I, Ruprich J (2001). The development of food consumption in the Czech Republic after 1989. Public Health Nutr 4, 999–1003.

  10. EFSA (2009). General principles for the collection of national food consumption data in the view of a pan-European dietary survey. EFSA J 7, 1–51.

  11. Field AP (2005). Is the meta-analysis of correlation coefficients accurate when population correlations vary? Psychol Methods 10, 444–467.

  12. Freedman LS, Kipnis V, Schatzkin A, Tasevska N, Potischman N (2010). Can we use biomarkers in combination with self-reports to strengthen the analysis of nutritional epidemiologic studies? Epidemiol Perspect Innov 7, 2.

  13. Galan P, Viteri FE, Bertrais S, Czernichow S, Faure H, Arnaud J et al. (2005). Serum concentrations of beta-carotene, vitamins C and E, zinc and selenium are influenced by sex, age, diet, smoking status, alcohol consumption and corpulence in a general French adult population. Eur J Clin Nutr 59, 1181–1190.

  14. German Institute of Human Nutrition (2009). The Multiple Source Method (MSM).

  15. Gordon HT, Bauernfeind JC (1982). Carotenoids as food colorants. Crit Rev Food Sci Nutr 18, 59–97.

  16. Hjartaker A, Lund E, Bjerve KS (1997). Serum phospholipid fatty acid composition and habitual intake of marine foods registered by a semi-quantitative food frequency questionnaire. Eur J Clin Nutr 51, 736–742.

  17. Ireland J, van Erp-Baart AM, Charrondiere UR, Moller A, Smithers G, Trichopoulou A (2002). Selection of a food classification system and a food composition database for future food consumption surveys. Eur J Clin Nutr 56 (Suppl 2), S33–S45.

  18. Jansen MC, Van Kappel AL, Ocke MC, Van ‘t Veer P, Boshuizen HC, Riboli E et al. (2004). Plasma carotenoid levels in Dutch men and women, and the relation with vegetable and fruit consumption. Eur J Clin Nutr 58, 1386–1395.

  19. Jenab M, Slimani N, Bictash M, Ferrari P, Bingham SA (2009). Biomarkers in nutritional epidemiology: applications, needs and new horizons. Hum Genet 125, 507–525.

  20. Jones ST, Aryana KJ, Losso JN (2005). Storage stability of lutein during ripening of cheddar cheese. J Dairy Sci 88, 1661–1670.

  21. Kaaks R, Riboli E, Sinha R (1997). Biochemical markers of dietary intake. IARC Sci Publ 142, 103–126.

  22. Kaluzny MA, Duncan LA, Merritt MV, Epps DE (1985). Rapid separation of lipid classes in high yield and purity using bonded phase columns. J Lipid Res 26, 135–140.

  23. Katan MB, Deslypere JP, van Birgelen AP, Penders M, Zegwaard M (1997). Kinetics of the incorporation of dietary fatty acids into serum cholesteryl esters, erythrocyte membranes, and adipose tissue: an 18-month controlled study. J Lipid Res 38, 2012–2022.

  24. Khan NC, West CE, de Pee S, Bosch D, Phuong HD, Hulshof PJ et al. (2007). The contribution of plant foods to the vitamin A supply of lactating women in Vietnam: a randomized controlled trial. Am J Clin Nutr 85, 1112–1120.

  25. Kleinbaum DG, Kupper LL, Muller KE, Nizam A (2007). Applied Regression Analysis and Other Multivariable Methods. Duxbury Press: Pacific Grove.

  26. Leaf DA, Connor WE, Barstad L, Sexton G (1995). Incorporation of dietary n-3 fatty acids into the fatty acids of human adipose tissue and plasma lipid classes. Am J Clin Nutr 62, 68–73.

  27. Macdonald HM, Hardcastle AC, Duthie GG, Duthie SJ, Aucott L, Sandison R et al. (2009). Changes in vitamin biomarkers during a 2-year intervention trial involving increased fruit and vegetable consumption by free-living volunteers. Br J Nutr 102, 1477–1486.

  28. Maiani G, Caston MJ, Catasta G, Toti E, Cambrodon IG, Bysted A et al. (2009). Carotenoids: actual knowledge on food sources, intakes, stability and bioavailability and their protective role in humans. Mol Nutr Food Res 53 (Suppl 2), S194–S218.

  29. Metcalfe LD, Schmitz AA, Pelka JR (1966). Rapid preparation of fatty acid esters from lipids for gas chromatographic analysis. Ana Chem 38, 514–515.

  30. Northrop-Clewes CA, Thurnham DI (2007). Monitoring micronutrients in cigarette smokers. Clin Chim Acta 377, 14–38.

  31. Pawlosky RJ, Hibbeln JR, Salem Jr N (2007). Compartmental analyses of plasma n-3 essential fatty acids among male and female smokers and nonsmokers. J Lipid Res 48, 935–943.

  32. Polsinelli ML, Rock CL, Henderson SA, Drewnowski A (1998). Plasma carotenoids as biomarkers of fruit and vegetable servings in women. J Am Diet Assoc 98, 194–196.

  33. Resnicow K, Odom E, Wang T, Dudley WN, Mitchell D, Vaughan R et al. (2000). Validation of three food frequency questionnaires and 24-h recalls with serum carotenoid levels in a sample of African-American adults. Am J Epidemiol 152, 1072–1080.

  34. Rock CL, Thornquist MD, Kristal AR, Patterson RE, Cooper DA, Neuhouser ML et al. (1999). Demographic, dietary and lifestyle factors differentially explain variability in serum carotenoids and fat-soluble vitamins: baseline results from the sentinel site of the Olestra Post-Marketing Surveillance Study. J Nutr 129, 855–864.

  35. Saadatian-Elahi M, Slimani N, Chajes V, Jenab M, Goudable J, Biessy C et al. (2009). 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. Am J Clin Nutr 89, 331–346.

  36. Simon JA, Fong J, Bernert Jr JT, Browner WS (1996). Relation of smoking and alcohol consumption to serum fatty acids. Am J Epidemiol 144, 325–334.

  37. Skeie G, Braaten T, Hjartaker A, Lentjes M, Amiano P, Jakszyn P et al. (2009). Use of dietary supplements in the European Prospective Investigation into Cancer and Nutrition calibration study. Eur J Clin Nutr 63 (Suppl 4), S226–S238.

  38. Slimani N, Deharveng G, Charrondiere RU, van Kappel AL, Ocke MC, Welch A et al. (1999). 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 58, 251–266.

  39. Slimani N, Ferrari P, Ocke M, Welch A, Boeing H, Liere M et al. (2000). Standardization of the 24-h diet recall calibration method used in the European prospective investigation into cancer and nutrition (EPIC): general concepts and preliminary results. Eur J Clin Nutr 54, 900–917.

  40. Slimani N, Valsta L (2002). Perspectives of using the EPIC-Soft programme in the context of pan-European nutritional monitoring surveys: methodological and practical implications. Eur J Clin Nutr 56 (Suppl 2), S63–S74.

  41. Souverein OW, Dekkers AL, Geelen A, Haubrock J, de Vries JH, Ocké MC et al. (2011). Comparing four methods to estimate usual intake distributions. Eur J Clin Nutr 65 (Suppl 1), S92–S101.

  42. Steingrimsdottir L, Ovesen L, Moreiras O, Jacob S (2002). Selection of relevant dietary indicators for health. Eur J Clin Nutr 56 (Suppl 2), S8–S11.

  43. Subar AF, Dodd KW, Guenther PM, Kipnis V, Midthune D, McDowell M et al. (2006). The food propensity questionnaire: concept, development, and validation for use as a covariate in a model to estimate usual food intake. J Am Diet Assoc 106, 1556–1563.

  44. Welch AA, Bingham SA, Ive J, Friesen MD, Wareham NJ, Riboli E et al. (2006). 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 84, 1330–1339.

  45. Yeum KJ, Booth SL, Sadowski JA, Liu C, Tang G, Krinsky NI et al. (1996). Human plasma carotenoid response to the ingestion of controlled diets high in fruits and vegetables. Am J Clin Nutr 64, 594–602.

  46. Zock PL, Mensink RP, Harryvan J, de Vries JH, Katan MB (1997). Fatty acids in serum cholesteryl esters as quantitative biomarkers of dietary intake in humans. Am J Epidemiol 145, 1114–1122.

Download references

Acknowledgements

This work was supported by the European Community funding under the Sixth Framework Program for the EFCOVAL Project (FOOD-CT-2006-022895). This document reflects only our views and the European Community is not liable for any use that may be made of the information contained therein. We thank all participants in this study.

Author information

Correspondence to S P Crispim.

Ethics declarations

Competing interests

JR received consulting fees from the Czech Technology Platform for food and healthy lifestyle. The remaining authors declare no conflict of interest.

Additional information

Contributors: SPC carried out data analyses and wrote the paper, taking into account comments from all co-authors. JHdV, AG and PvV designed and coordinated the validation study. OWS contributed to the statistical analyses. PJMH was responsible for laboratorial analyses. JHdV, AG, LL, A-SR, ITL, LFA, IH, WDK, JR, MD, MCO and NS were involved in the fieldwork and gave input on interpretation of results. All co-authors commented on the paper and approved the final version.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • fish
  • fruits
  • vegetables
  • biomarker
  • validity
  • 24-h dietary recall

Further reading