Introduction
Nutritional epidemiological research requires addressing issues of measurement errors and inter and intra-individual variability, which are specific for each nutrient. Public health decisions must rely on valid and precise estimates of nutrient intake and status. There is a need to reach a consensus about the best available methods for assessing nutrient intake and status at the population level. In this article a literature review of dietary methods used to assess intake of n-3 PUFA is presented. Also biomarkers for n-3 PUFA status were analysed. The effect of dietary fats on health and disease has been of interest for many decades. The various health benefits of consuming the LC n-3 PUFA particularly eicosapentaenoic acid (EPA, 20 : 5n-3) and docosapenteanoic acid (DHA, 22 : 6n-3), have been widely reported(Reference Sullivan, Williams and Meyer1). The LC n-3 PUFA are obtained predominantly from fish, seafood, meat, and eggs(Reference Sullivan, Williams and Meyer1). However, various dietary supplements containing several hundred milligrams of LC n-3 PUFAs per dose are commonly available. Many clinical studies have assessed the effect of LC n-3 PUFA supplementation in restoring health and maintaining well-being. The majority of these reviews concluded that, although there was some indication of the beneficial effect of LC n-3 PUFA supplementation, further studies were needed to establish efficacy of their use. To date, there is lack of a universally accepted biomarker that reflects increased LC n-3 PUFA status in response to increased dietary intake or supplementation. It is even more important in epidemiologic studies assessing health effects of LC n-3 PUFA status in populations over a long-term period to understand which biomarkers truly reflect LC n-3 PUFA status. To assess the reliability of biomarkers in reflecting LC n-3 PUFA intake, it is necessary to review biomarker data from studies reporting a change in LC n-3 PUFA status. On the other hand it is also necessary to know the validity and reproducibility of dietary intake estimations of LC n-3 PUFA from different questionnaires with regard to the appropiatte biomarkers. Therefore, the aims of this paper were to review the validity of methods used to measure the usual n-3 PUFA intake of a population and additionally, to assess the usefulness of different biomarkers of LC n-3 PUFA status in healthy humans.
Methods
This article includes two updated systematic reviews. Both systematic literature searches were performed between March and May 2011. Two previous systematic reviews covering the objectives of both searches were conducted in 2007 and 2009 within the European Network of Excellence EURopean micronutrient RECommendations Aligned (EURRECA)(Reference Ashwell, Lambert and Alles2).
For the first search updating the validity of methods to assess usual n-3 PUFA intake, the literature search was conducted in Medline, OvidSP and EMBASE using the following terms: ‘omega-3 fatty acid’, ‘fish oils’, ‘biomarker’, ‘nutritional assessment’, and ‘fat intake’ including MESH-terms. In total 286 articles were selected using Medline, 358 were selected from OvidSP and 330 were identified from EMBASE.
To select the articles to be included in the present review the following exclusion criteria were used: (a) studies conducted exclusively in diseased or institutionalised persons, (b) studies relating diseases to food consumption or nutrient intake, (c) intervention studies and other therapeutic studies with nutrients or drugs related to the metabolism of these nutrients, (d) studies in animals, (e) studies written in languages other than English or Spanish, (f) studies using single 24-hour recall or non validated FFQ, (g) studies related to fish consumption, (h) studies in infants and children, and (i) studies using another dietary method as a reference tool.
Nine hundred seventy four titles and abstracts were identified via the electronic search from the three different databases. After excluding duplicate studies, a total of 87 appeared to be potentially relevant, and we attempted to obtain them in full-text version. The literature lists in the selected papers were checked and consequently 3 more articles were included. From the 90 articles, 8 of them were chosen to update Table 1 elaborated in the original article from 2009(Reference Øverby, Serra-Majem and Frost Andersen3). In total 19 studies were reviewed (11 were already in the first review and 8 were consequently added). To assess the quality of the different calibration/validation studies a quality score system was developed(Reference Serra-Majem, Frost Andersen and Henriquez-Sánchez4). This has been described in previous publications by Serra-Majem et al. (Reference Serra-Majem, Frost Andersen and Henriquez-Sánchez4) and Øverby et al. (Reference Øverby, Serra-Majem and Frost Andersen3).
Table 1 Description of the studies included in this review validating intake of n-3 fatty acids (sorted by publication date). Crude and adjusted correlations for dietary methods vs. reference methods in the studies included
DHQ, diet history questionnaire; ALA, α-linolenic acid; DPA, docosapentanoic acid.
Significance: * P < 0·05; ** P < 0·01; *** P < 0·001; **** P < 0·0001.
† Spearman correlation.
‡ Marine intake of n-3 FA.
§ Pearson correlation.
∥ Deattenuated with the within-to-between person variance ratio for intake of FA.
¶ Corrected after attenuation correction factor.
†† Corrected for the reliability coefficients of FFQ and phospholipids.
§§ Adjusted for age at blood drawing, BMI, current weight, smoking status, postmenopausal status, postmenopausal hormone use, period of blood assay, and fasting status at blood drawing.
∥∥ Adjusted for age and total energy intake.
For the second search of the present article aiming to assess the utility of biomarkers for n-3 PUFA, another OvidSP and MEDLINE search was developed in order to refresh the search strategy developed by Fekete et al. (Reference Fekete, Marosvölgyi and Jakobik5) in the systematic review of recovery studies. In this case, the search targeted intervention/recovery studies of n-3 PUFA using text terms with appropriate truncation and relevant indexing terms. The following strategy was applied: (n-3 LCPUFA terms) and (intervention study terms) and (human studies) and was limited to the last 4 years. The inclusion criteria was the same as that used by Fekete et al. (2007)(Reference Fekete, Marosvölgyi and Jakobik5). Six hundred and twenty one titles and abstracts were identified via electronic search. From these, only 8 studies were selected to update studies included in the initial study conducted by Fekete et al. (Reference Fekete, Marosvölgyi and Jakobik5)
Biomarkers
Eighteen different biomarkers were used to characterize changes in LC n-3 PUFA status. Discussion is only included for those biomarkers used in more than 3 different studies. Data for each study included in the present analysis is described in Table 2. The effects of LC n-3 PUFA supplementation on each biomarker are detailed in Table 3. The main focus is directed towards the effect of DHA supplementation on biomarkers reflecting changes in DHA values. Plasma phospholipid DHA as well as erythrocyte and platelet DHA appear to be reliable and robust biomarkers as shown in Table 3.
Table 2 Basic characteristics of the included studies on recovery interventions for n-3 PUFA biomarkers1
1 M, exclusively male group; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; RCT, randomized controlled trial; p, parallel; PPL, plasma phospholipids; X, mixed sex group; ALA, a-linolenic acid; B/A, before-after study; Plat, total platelets; G, total granulocyte; NPL, neutrophil phospholipid; NEFA, nonesterified fatty acids; E, total erythrocytes; F, exclusively female group; EPL, erythrocyte phospholipids; P, total plasma; PCE, plasma cholesteryl esters; PTG, plasma triacylglycerols; PL, phospholipids; N, total neutrophils; DPA, docosapentanoic acid; PBMC, peripheral blood mononuclear cells
Table 3 Primary analyses (pooled data on the longest duration and the highest supplementation dose) for each identified biomarker1
Modified and updated from Fekete et al. (Reference Fekete, Marosvölgyi and Jakobik5).
1 WMD, weighted mean difference; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; N/A, no available data; PBMC, peripheral blood mononuclear cell.
2 All studies are in %DHA of total fatty acids unless otherwise stated.
3 To claim that a biomarker was effective (reflected change in status) within a review, 3 conditions needed to be met: (1) statistical significance within a forest plot (95 % CI did not include 0 or P < 0·05); (2) ≥ 3 trials contributed data; and (3) between intervention and control arms in the studies contributing data there were ≥ 50 participants. To claim that a biomarker was ineffective, 4 conditions had to be met: (1) lack of statistical significance within a forest plot (95 % CI included 0 or P ≥ 0·05); (2) ≥ 3 trials contributed data; (3) between intervention and control arms in the studies contributing data there were ≥ 50 participants; and (4) study results were approximately similar (heterogeneity levels were acceptable so that I 2 < 50 %).
4 %EPA of total fatty acid.
5 μg/mg protein.
Results
Dietary method studies
Details of the 19 papers selected are given in Table 1. In the 19 articles included in the review, 15 different food frequency questionnaires (FFQ) were validated. All FFQs were designed to capture the usual diet. Some questionnaires specifically asked only about n-3-PUFA rich food(Reference Sullivan, Williams and Meyer1, Reference Sublette, Segal-Isaacson and Cooper6), while others covered the whole diet with 66–360 food items included in the questionnaire(Reference Hunter, Rimm and Sacks7–Reference Ma, Folsom and Shahar17). A diet history questionnaire had been validated in one study(Reference Sasaki, Ushio and Amano18). Weighed records had been validated in 4 studies(Reference McNaughton, Hughes and Marks12, Reference Marckmann, Lassen and Haraldsdottir19, Reference Kobayashi, Sasaki and Kawabata20, Reference Kuriki, Nagaya and Tokudome21).
In the presented studies the numbers of participants varied from 24 to 4439. The age distribution ranged from 18 to 86 years, with mean ages from 45 to 65 years. In total 15 different FFQs and dietary records or recalls in 5 different settings (varying number of days and season) were validated against subcutaneous fat, serum or plasma fatty acids.
Subcutaneous fat
Adipose tissue fatty acids were determined using chromatography and calculating the area under the curve for each of the fatty acids. All studies using fatty acids in tissue reported the same procedure with only slight modifications(Reference Hunter, Rimm and Sacks7, Reference Andersen, Solvoll and Johansson8, Reference Godley, Campbell and Miller10, Reference Knutsen, Fraser and Beeson13, Reference Baylin, Kabagambe and Siles14, Reference Marckmann, Lassen and Haraldsdottir19).
Five different FFQ were validated against adipose tissue(Reference Hunter, Rimm and Sacks7, Reference Andersen, Solvoll and Johansson8, Reference Godley, Campbell and Miller10, Reference Knutsen, Fraser and Beeson13, Reference Baylin, Kabagambe and Siles14). All these correlations were significant. Furthermore Marckmann et al. (Reference Marckmann, Lassen and Haraldsdottir19) validated weighed records (3 × 7 d) against subcutaneous fat. Only DHA crude correlations were significant. Finally Knutsen et al. (Reference Knutsen, Fraser and Beeson13) validated eight different 24-h recalls of intake of ALA, EPA and DHA against subcutaneous fat. They found high adjusted correlations for ALA, while the correlations for EPA and DHA were lower (Table 1).
Blood component concentrations
After extraction and isolation the serum/plasma phospholipids were quantified by gas liquid chromatography after methylation(Reference Sublette, Segal-Isaacson and Cooper6, Reference Andersen, Solvoll and Johansson8, Reference Arsenault, Matthan and Scott9, Reference Hodge, Simpson and Gibson11, Reference McNaughton, Hughes and Marks12, Reference Sun, Ma and Campos15–Reference Sasaki, Ushio and Amano18, Reference Kobayashi, Sasaki and Kawabata20–Reference Astorg, Bertrais and Laporte23). Some expressed the serum phospholipids as mg fatty acid/l serum(Reference Hjartaker, Lund and Bjerve22), while most used percent of total fatty acid methyl esters(Reference Andersen, Solvoll and Johansson8, Reference Sasaki, Ushio and Amano18, Reference Kobayashi, Sasaki and Kawabata20) or both(Reference Hodge, Simpson and Gibson11). For detailed descriptions refer to each particular study.
Eleven different FFQ were validated against erythrocytes, plasma or serum(Reference Sullivan, Williams and Meyer1, Reference Sublette, Segal-Isaacson and Cooper6, Reference Andersen, Solvoll and Johansson8, Reference Arsenault, Matthan and Scott9, Reference Hodge, Simpson and Gibson11, Reference McNaughton, Hughes and Marks12, Reference Sun, Ma and Campos15–Reference Ma, Folsom and Shahar17, Reference Hjartaker, Lund and Bjerve22, Reference Livingstone, Prentice and Strain24). Sullivan et al. (Reference Sullivan, Williams and Meyer1) validated fatty acid estimated from a FFQ against both fatty acid from erythrocytes and from plasma. All the correlations were significant. Andersen et al. (Reference Andersen, Solvoll and Johansson8), Hjartaker et al. (Reference Hjartaker, Lund and Bjerve22) and Hodge et al. (Reference Hodge, Simpson and Gibson11) reported significant correlations of approximately 0·50–0·60 between dietary intake of EPA and DHA estimated from the FFQ and concentrations of EPA and DHA in serum or plasma. Arsenault et al. (Reference Arsenault, Matthan and Scott9) reported adjusted correlations between dietary intake of fatty acids estimated from the FFQ in controls and concentrations of fatty acid in plasma of 0·38 for EPA and 0·49 for DHA. Godley et al. (Reference Godley, Campbell and Miller10) reported correlations between dietary intake of EPA and DHA estimated from the FFQ and concentrations of EPA and DHA in erythrocyte membrane ranging from 0·19 to 0·36. This is the smallest value found for DHA in the blood biomarkers. Sublette et al. (Reference Sublette, Segal-Isaacson and Cooper6) reported significant correlations between dietary intake of ALA, EPA and DHA estimated from the FFQ and concentrations of ALA, EPA and DHA in plasma of 0·22 for ALA, 0·38 for EPA and 0·50 for DHA, which was the highest value found for the different biomarkers utilized. Sun et al. (Reference Sun, Ma and Campos15) reported a significant adjusted correlation of 0·56 between dietary intake of DHA estimated from the FFQ and concentrations of DHA in erythrocytes.
One dietary history questionnaire was also validated against serum fatty acid and high crude (r 0·46) and adjusted (r 0·59) correlations were reported for intake of EPA for men(Reference Sasaki, Ushio and Amano18). This questionnaire was self-administered and was somewhat similar to a FFQ.
Three studies have validated weighed records (all with seven or more days) against serum, erythrocytes or plasma fatty acids(Reference McNaughton, Hughes and Marks12, Reference Kobayashi, Sasaki and Kawabata20, Reference Kuriki, Nagaya and Tokudome21). Kobayashi et al. (Reference Kobayashi, Sasaki and Kawabata20) presented a very high correlation coefficient for EPA, crude (r 0·75) and adjusted (r 0·89), as well as the best adjusted correlations for DHA and total n-3 PUFA from weighed records validated against serum fatty acids (r 0·61). Kuriki et al. (Reference Kuriki, Nagaya and Tokudome21) obtained adjusted correlations for dietary intake of EPA measured with weighed records against plasma concentrations of EPA (r 0·57) and for DHA (r 0·57). McNaughton et al. (Reference McNaughton, Hughes and Marks12) showed a crude correlation of 0·43 for DHA measured with weighed records validated against DHA concentration in plasma and a lower correlation coefficient for EPA (r 0·22). Similar correlations were observed when the intake was measured with a FFQ (DHA r 0·32 and EPA r 0·21). All three studies presented low correlations for ALA(Reference McNaughton, Hughes and Marks12, Reference Kobayashi, Sasaki and Kawabata20, Reference Kuriki, Nagaya and Tokudome21).
Biomarker study
A total of 8 new studies were incorporated to the 41 previously included papers identified in the study by Fekete et al. (Reference Fekete, Marosvölgyi and Jakobik5).
Details of the biomarkers analysed are given in Table 2 wich include 49 recovery studies of n-3 PUFA biomarkers. In this study, as summarized in Table 3, Total plasma lipid DHA appears to be a good biomarker of DHA status, which reacts rapidly to supplementation and is sensitive to supplementation dose. It appears to be reliable in adults, mixed sex studies, and those with moderate baseline DHA status, but it is not clear for which other population subgroups its application can be reliable. Moreover, plasma phospholipid DHA appears to be a good biomarker of DHA status. It reacts rapidly to supplementation and is also sensitive to supplementation dose. This biomarker appears to respond appropriately in adults, males, females, those with low, moderate, or high baseline DHA status, those who used marine oil, seafood, or single cell oils, and in those whose dose amounts were ≤ 2500 mg/d of DHA. There were insufficient studies to assess the effectiveness of plasma phospholipid DHA in other population subgroups. With reference to Plasma phospholipid EPA, it appeared to be a good biomarker for EPA status in men and women and those who had low, moderate, or high baseline EPA, which reacts rapidly to supplementation and is sensitive to supplementation dose.
Plasma triacylglycerol DHA could be a good biomarker of DHA status, but there were insufficient studies to allow exploration of which population groups it may be most effective in. Plasma cholesteryl ester DHA appears to be a good biomarker of DHA status at lower-dose supplementation, but it is not clear within which population groups it is effective or whether it works well at higher doses of supplementation. Plasma nonesterified fatty acid DHA may be a good biomarker of DHA status, but there were insufficient studies to allow for the exploration of appropriateness of its use in different population groups and doses. Erythrocyte membrane total lipid DH appears to be a good biomarker of DHA status, and the data suggest that there is a dose response. Although it seems to be an effective biomarker in infants for most doses, confirmation is not possible due to limited data. Erythrocyte membrane phospholipid DHA appears to be a good biomarker of DHA status and although it seems to be an effective biomarker in adults, children and adolescents, as well as in pregnant or lactating women and at most doses, this cannot be confirmed due to limited data. Total platelet lipid DHA could be a good biomarker of DHA status, but there was no apparent dose response. For Peripheral blood mononuclear cell phospholipid DHA, the response of peripheral blood mononuclear cell phospholipid DHA values to DHA supplementation did not appear to be a good biomarker of DHA status. For other potential biomarkers of DHA status, evidence was insufficient for young erythrocyte ghost DHA, old erythrocyte ghost DHA, granulocyte DHA, neutrophil DHA, neutrophil phospholipid DHA, peripheral blood mononuclear cell total lipid DHA, LDL DHA, and HDL phospholipid DHA.
Discussion
In a validation study, the reference method used should be as accurate as possible(Reference Andersen, Solvoll and Johansson8). A validation study is also called a relative validation/calibration study when one dietary method is compared to another dietary method, most often FFQ v. several days of food records. The limitations with this approach are the considerable individual day-to-day variation, which reduces the possibility of obtaining a true measure of usual intake with few recording days, as well as reporting bias since both self-administered dietary assessment questionnaires and dietary records are based on self-reporting(Reference Andersen, Solvoll and Johansson8). FFQs often overestimate intake of energy and nutrients, while food records often underreport energy and nutrient intakes(Reference Livingstone, Prentice and Strain24, Reference Black, Goldberg and Jebb25). As such, we thought it best to exclude those questionnaires where validation was made against another dietary measurement tool. An alternative to relative validations is the use of biomarkers, whose primary advantage is that these measurements are objective and the sources of errors for a biomarker and a dietary assessment method are independent(Reference Andersen, Solvoll and Johansson8, Reference Arab26). The n-3 PUFA are largely exogenic, meaning that there is no synthesis of n-3 PUFA in the body and that intake via diet and supplements are the major source, making the correlations with biomarkers easier(Reference Hodge, Simpson and Gibson11, Reference Arterburn, Hall and Oken27). There are several choices of a biomarker for the measurement of LC n-3 PUFA, and those presented in this review were fatty acids in adipose tissue, erythrocytes and plasma. Adipose tissue fatty acids are generally considered the best source of assessing long-term fatty acid intake(Reference Knutsen, Fraser and Beeson13, Reference Arterburn, Hall and Oken27). Erythrocytes may be a useful marker as they can provide an indication of the previous 120-d intake of LC n-3 PUFA(Reference Sullivan, Williams and Meyer1). Plasma fatty acids reflect intake of fatty acids over the past few days or more(Reference Hodge, Simpson and Gibson11). Most of the included studies have presented the correlations, both crude and adjusted. The correlation coefficients obtained from the validation studies can reflect the capability of the method to rank individuals according to fatty acid intake.
Subcutaneous fat
Fatty acids estimated from six different FFQ(Reference Hunter, Rimm and Sacks7, Reference Andersen, Solvoll and Johansson8, Reference Godley, Campbell and Miller10, Reference Knutsen, Fraser and Beeson13, Reference Baylin, Kabagambe and Siles14), one weighed record(Reference Marckmann, Lassen and Haraldsdottir19) and one recall(Reference Knutsen, Fraser and Beeson13) were validated against subcutaneous fat, which the literature describes as the best reference method. The correlation coefficients observed in all the studies were in the range of 0·40–0·66 for ALA, EPA and DHA. In summary, none of the dietary methods validated against subcutaneous fat and presented here seem to be superior than the others in relation to ranking the dietary intake of n-3 PUFA. Two articles related to subcutaneous fat were found for this updated review, the correlation coefficient range for EPA was 0·15–0·33 and for DHA 0·18–0·42. This suggests a weaker correlation when compared to the previous correlation reported by Marckman et al. (Reference Marckmann, Lassen and Haraldsdottir19) which showed the highest correlation coefficient (0·66) amongst all the included studies.
Blood component composition
Dietary intake of n-3 PUFA estimated from eleven different FFQ(Reference Sullivan, Williams and Meyer1, Reference Sublette, Segal-Isaacson and Cooper6, Reference Andersen, Solvoll and Johansson8–Reference McNaughton, Hughes and Marks12, Reference Sun, Ma and Campos15–Reference Ma, Folsom and Shahar17, Reference Hjartaker, Lund and Bjerve22), one diet history questionnaire(Reference Sasaki, Ushio and Amano18) and three weighed record studies(Reference McNaughton, Hughes and Marks12, Reference Kobayashi, Sasaki and Kawabata20, Reference Kuriki, Nagaya and Tokudome21) was validated against fatty acids in serum, plasma or erythrocytes. Both fatty acids in plasma, erythrocytes and serum were found to be good biomarkers of LC n-3 PUFA(Reference Sullivan, Williams and Meyer1, Reference Hjartaker, Lund and Bjerve22). The correlation coefficients observed between the intake of fatty acids measured by most FFQs(Reference Sullivan, Williams and Meyer1, Reference Hunter, Rimm and Sacks7, Reference Andersen, Solvoll and Johansson8), the diet history questionnaire(Reference Sullivan, Williams and Meyer1), and the weighed records(Reference Kuriki, Nagaya and Tokudome21) v. fatty acid in blood parameters were at the same range (r 0·40–0·60). The best correlation was observed in the study by Kobayashi et al. (Reference Kobayashi, Sasaki and Kawabata20) comparing the dietary intake of fatty acids from weighed records with fatty acids in serum phospholipids (EPA, r 0·89). However, there was no clear tendency among the three studies comparing fatty acids from weighed records with fatty acids in blood(Reference McNaughton, Hughes and Marks12, Reference Kobayashi, Sasaki and Kawabata20, Reference Kuriki, Nagaya and Tokudome21). As such, it seems that weighed records were the best method to measure n-3 PUFA intake.
Most correlation coefficients from the studies comparing dietary intake with fatty acids in blood parameters were in the same range as the ones observed for fatty acids in adipose tissue (r 0·40–0·60). There were two studies with a lower correlation(Reference McNaughton, Hughes and Marks12, Reference Astorg, Bertrais and Laporte23) and one with a correlation higher than this range(Reference Kobayashi, Sasaki and Kawabata20). For ALA most studies presented low correlations between dietary intake and blood parameters in both previous and updated versions of this review(Reference Sublette, Segal-Isaacson and Cooper6, Reference Sun, Ma and Campos15, Reference Ma, Folsom and Shahar17, Reference Astorg, Bertrais and Laporte23). In the present updated review, the correlation coefficient range for erythrocyte EPA was 0·23–0·38 and for erythrocyte DHA 0·19–0·56 which suggests an acceptable and a reasonable good correlation coefficient respectively(Reference Godley, Campbell and Miller10, Reference Sun, Ma and Campos15). Regarding plasma phospholipid EPA the correlation range was 0·21–0·38 and for DHA 0·25–0·50(Reference Sublette, Segal-Isaacson and Cooper6, Reference Sun, Ma and Campos15, Reference Astorg, Bertrais and Laporte23), again showing an acceptable and reasonably good correlation coefficient, respectively. It is important to highlight that none of the current correlation coefficients found in this updated review were higher than those previously reported in the original article; however the same levels were maintained. This implies that any additional validation study will unlikely produce higher correlation estimates between questionnaires and biomarkers.
The estimation of summarised crude and adjusted correlations for all the validation studies of FFQs using biomarkers as the reference method indicates that the FFQ gives ‘acceptable’ values for total n-3 PUFA, EPA and DHA. The summarised crude and adjusted correlations for the two studies validating weighed records against biomarkers indicate ‘acceptable’ estimates for total n-3 PUFA, while the estimates obtained a higher ranking of ‘good’ for EPA and DHA. As expected, the weighed records seem to be superior to the FFQ in reference to estimating intakes of EPA and DHA.
Biomarkers were more accurate than different dietary methods to rank individuals. One limitation with food records is that subjects are prone to underestimate their food intake when they keep food records(Reference Livingstone, Prentice and Strain24). The true food consumption of n-3 FA most likely lies somewhere between the weighed records and the FFQ. According to the systematic review, none of the dietary assessment methods used to assess n-3 PUFA seem to be highly superior to another, with weighed records being slightly better than FFQs. Most studies presented correlation coefficients ranging from 0·40 to 0·60. This review also confirmed the view that employing an FFQ to assess n-3 PUFA requires that it be validated against reliable and valid biomarkers, and that validation studies of dietary methods for measuring intakes of n-3 FA could be improved.
Additionally, after analyzing the 18 different potential biomarkers of LC n-3 PUFA status reported in Table 3, we could argue that plasma phospholipid DHA, erythrocyte DHA and platelet DHA were all effective biomarkers of DHA status. With regard to other biomarkers (plasma DHA, plasma triacylglycerol DHA, plasma cholesteryl ester DHA, plasma nonesterified DHA, erythrocyte phospholipid DHA, and plasma phospholipid EPA) we could not find any additional evidence to support modifying or promoting previously published statements from Fekete et al. (Reference Fekete, Marosvölgyi and Jakobik5).
Still and all, it is worth noting that the health benefits of increasing LC n-3 PUFA dietary intake need to be evaluated in RCTs investigating specific clinical outcomes. There are some clear limitations found. First, the number of studies reporting data on different potential biomarkers is limited. This situation reduced our ability to explore which population subgroups or in which types of intervention the biomarkers are effective. Second, we were able to focus on the effect of supplementing DHA only, whereas LC n-3 PUFA supplementation usually consists of a complex mixture of LC n-3 PUFAs that may interconvert with each other(Reference Arterburn, Hall and Oken27). Third, the dose-response curve of the incorporation of DHA (or any other fatty acid) may differ between the distinct blood component constituents(Reference Brown, Pang and Roberts28, Reference Tremoli, Maderna and Marangoni29); hence, it may be assumed, with good reason, that the uniform time duration and dose categories may have differently influenced the evaluation of the biomarkers.
Although several clinical studies have investigated the response of various biomarkers to modified n-3 fatty acid intake(Reference Brown, Pang and Roberts28–Reference Cao, Schwichtenberg and Hanson33) and important theoretical considerations have also been published(Reference Harris, Sands and Windsor34, Reference Fokkema, Smit and Martini35, Reference Harris, Assaad and Poston36), we still do not have enough data available in the literature. Which biomarker might be sensitive enough to detect changes of a given dose of LC n-3 PUFA supplementation in a given clinical condition or population group? Further research is needed to characterize and to understand the meaning of the different correlations between intake estimates and biomarkers of LC n3 PUFA in distinct population groups and environments.
Acknowledgements and disclosures
The preparatory meetings for this series of reviews on fat and health were funded by Puleva Food. Neither Lluis Serra-Majem nor Mariela Nissensohn, Nina C Øverby or Katalin Fekete have conflicts of interest to disclose. Lluis Serra-Majem and Mariela Nissensohn contributed to the design of the strategy for the literature search, double screened and selected the retrieved documents. Authors acknowledge Alexandra Santos Johnson and Melania Morales Sanchez from de Department of Clinical Sciences of the University of Las Palmas de Gran Canaria (ULPGC) the support provided to select and retrieve the documents, as well as Francisco Fumagallo from the Library of the School of Health Sciences at the ULPGC. Nina C Øverby and Katalin Fekete provided previous literature searches and analysis. Lluis Serra-Majem prepared the main outline of the manuscript and all authors contributed to the preparation of the manuscript.
