Results from this study demonstrate that the FFQ appears to provide adequate information on n-3 intake in relation to corresponding FA levels in plasma PL. In particular, dietary intakes estimated by a FFQ were correlated with percentages of DHA and total n-3 FA in plasma PL.
FFQ estimates of dietary intakes for EPA, DPA and DHA in this young and well-educated French-Canadian population were comparable to results reported in previous studies [27–33]. Yet, ALA and total n-3 FA intakes were higher compared to intakes observed in earlier studies [27–33]. The estimated daily dietary intake of ALA in most European countries and United States is of 1.3 to 1.7 g [34, 35]. Thus, the higher intake of ALA may be due to the type of vegetable oil used by this population and thus may have beneficial effects on the cardiovascular disease risk profile. Yet, the consumption of EPA and DHA in this young and well-educated French-Canadian study population was below the level of 0.5 g/day recommended to decrease the risk of cardiovascular disease [36, 37]. The combined daily intake of EPA and DHA was 0.25 g for men and 0.27 g for women. Similarly, Lucas et al., examined the n-3 FA status of individuals living in the Quebec City metropolitan area and established that the mean daily intake of EPA and DHA was of 0.29 g . However, in our study the n-3 FA intake was higher than that reported in the NHANES 1999–2000, where mean intake of EPA plus DHA was approximately 0.1 g/d . Nevertheless, the present intake of EPA and DHA is much lower than the 1.6 g daily intake observed in the Japanese population, which has one of the lowest rates of heart disease in the world .
Even though the FFQ estimates are comparable to other studies, percentages of n-3 FA in plasma PL of the participants were lower than the percentages previously reported in the literature [17, 30, 41]. In particular, plasma DHA and total n-3 percentage were much lower compared to the average from the Melbourne Collaborative Cohort Study (MCSS) (4.0-4.2% DHA and 6.6-6.8% total n-3) , the European Prospective Investigation into Cancer (EPIC) -Norfolk United Kingdom cohort (5.13-5.28% DHA and 7.81-7.96% total n-3 FA) , and the entire EPIC cohort (4.6-6.6% DHA, and 6.7-10.4% total n-3). According to several studies, FA composition stay stable for many years when plasma is stored at −80 °C [19, 21, 42]; therefore, the poor levels of DHA and total n-3 FA observed in the current study are unlikely to be explained by FA oxidation. Overall, results indicate lower levels of n-3 FA in plasma PL in the French-Canadian population.
Results show a modest relationship between FFQ estimates of dietary intakes and percentages of n-3 FA in plasma PL, except for ALA in women and ALA, EPA and DPA in men. The results of the present study are in agreement with results reported in previous studies [15, 27, 28, 32, 43, 44]. First, a validity study was conducted in Australia for a non-specific FFQ using plasma PL . The results demonstrate significant relationships only for DHA (0.32) and total long-chain n-3 (0.38) . Secondly, Hodge et al. , obtained similar results for ALA, EPA, DHA and total n-3 FA with a general FFQ (0.07, 0.18, 0.40 and 0.31, respectively). Thirdly, Arsenault et al.  demonstrated correlation coefficients of 0.37 for EPA, 0.48 for DHA and 0.48 for total n-3 FA between an n-3 PUFA specific FFQ and plasma PL among elderly with cognitive impairment. Finally, earlier studies conducted with other biomarkers, such as red blood cells, total plasma and adipose tissue, demonstrated comparable correlation coefficients [15, 43–50]. Further, these results suggest that the lack of correlation between FFQ and plasma PL for ALA may be due to the fact that ALA is readily metabolized to long-chain metabolites, such as EPA and DHA [47, 51] . Overall, the FFQ adequately estimated DHA and total n-3 FA in the population as well as EPA and DPA in women.
Taken together, results demonstrate modest correlations (0.23-0.57) between n-3 dietary intakes estimated using a FFQ and plasma PL. It is well accepted that perfect correlations between FA in tissues and intakes measured by FFQs are unrealistic. First, numerous factors can influence the FA composition such as the tissue used for FA profile as well as the individuals’ FA metabolism. Since the FFQ assesses the dietary intakes of the last month, plasma PL should be a reasonable biomarker to evaluate this information against as it reflects fatty acid intake over the last 21 d [27, 52]. Erythrocytes are considered the gold standard to measure the long-term n-3 status . Yet, numerous studies have reported strong correlations between FA content of erythrocytes and plasma PL; therefore, plasma PL may be considered as a reliable biomarker to reflect n-3 status over the last month [21, 24]. Furthermore, it is well recognized that FA concentrations in tissues are subjected to other issues such as absorption, metabolism, genetic and lifestyle determinants [14, 17, 18, 20]. Secondly, the FFQ data are influenced by the capacity of individuals to recall adequately food they consumed, portions they eat and by their knowledge of food’s composition [19, 21]. In addition, the FFQ estimates are also subjected to the accuracy of the databases used . In sum, the modest correlation coefficients observed in the current study support the ability of this FFQ to reflect the relative ranking of n-3 FA intake.
In conclusion, our results indicate intakes of approximately 0.25 to 0.27 g/d EPA and DHA in this young and educated French-Canadian population which is below the recommended 0.5 g/d EPA and DHA for cardiovascular disease risk reduction [34, 36, 37]. The present study also indicates that n-3 FA intakes examined by FFQ and validated with n-3 FA biomarkers, are good indicators of n-3 FA status. Taken as a whole, the FFQ could be used to categorize n-3 FA status, especially due to a lower cost compared to established biomarkers. Nevertheless, the use of the questionnaire with a less educated or older population may be less accurate compared to this particular French-Canadian population. Overall, the use of this FFQ could facilitate future studies in investigating the impact of low n-3 FA dietary intakes on health.