Very-long-chain ω-3 fatty acids as markers for habitual fish intake in a population consuming mainly lean fish: the EPIC cohort of Gipuzkoa

Abstract

Objective: To assess the relationship between habitual fish intake and fatty acid levels in serum as well as in the LDL fractions of serum phospholipids and cholesteryl esters. Design: Cross-sectional study. Setting: Cohort of Gipuzkoa (Basque Country, northern Spain) included in the European Prospective Investigation into Cancer and Nutrition (EPIC) project. Subjects: Random sample of 120 healthy volunteers of both sexes aged 35 -65 y, divided into various consumption groups according to daily fish intake. Methods: Data on habitual intake over the previous year was collected by trained interviewers by means of a computerized questionnaire based on the diet history method. Fasting venous blood samples were drawn and fatty acids were measured by gas-liquid chromatography. Results: Lean fish accounted for 78% of all fish consumption in the highest consumption group ( > 115 g=day) and for 60% in the lowest ( < 31 g=day). The mean concentrations of o-3 polyunsaturated fatty acids (PUFA), eicosapentaenoic acid (EPA, C20:5, o-3), and docosahexaenoic acid (DHA, C22:6, o-3) in serum and in the LDL fractions of serum phospholipids and cholesteryl esters increased significantly from the lowest to the highest fish consumption categories. Fish intake showed a statistically significant relationship with o-3 PUFA, EPA and DHA in serum and in the LDL fractions of serum phospholipids and cholesteryl esters both in the simple linear regression analysis and in a multiple regression model adjusted by age, body mass index (BMI) and vegetable intake. Conclusions: Habitual fish intake is reflected in the content of EPA and DHA in serum and in the LDL phospholipid and cholesteryl esters fractions. The concentrations of very-long-chain o-3 fatty acids are useful biomarkers for dietary fish intake, mainly lean fish. Sponsorship: Europe Against Cancer Programme of the European Union (agreement SOC 97 200302 05F02); 'Fondo de Investigaciones Sanitarias', Spanish Ministry of Health (FIS grant 99=0024-05); Government of the Basque Country; and 'Fundación Científica de la Asociación Española contra el Cáncer'. Descriptors: fish intake; fish oil; n-3 fatty acids; biological markers; eicosapentaenoic acid; docosahexaenoic acid

Plasma n-6 and n-3 polyunsaturated fatty acids as biomarkers of their dietary intakes: a cross-sectional study within a cohort of middle-aged French men and women

“…At variance, the plasma linoleic acid levels reach values (30-35% of fatty acids) comparable to those found in United Kingdom (Rosell et al, 2005), Norway (Andersen et al, 1999), Japan (Kuriki et al, 2002(Kuriki et al, , 2003 or Costa-Rica (Kuriki et al, 2003;Baylin et al, 2005), reflecting the high intakes of linoleic acid in most world countries (4-6% of energy) (Simopoulos, 2002). Plasma levels of long-chain n-3 PUFA in our sample are much higher than in populations eating no or little fish (Baylin et al, 2005;Rosell et al, 2005), similar to those found in Norway (Andersen et al, 1999), and lower than those found in countries with a high fish consumption such as Spain (Amiano et al, 2001) or Japan (Kuriki et al, 2002(Kuriki et al, , 2003. In Europe, the French rank among the higher fish and seafood consumers, after the Spanish and the Scandinavians (Welch et al, 2002).…”

“…In addition, specificities in the fatty acid composition of blood or tissue lipid classes make that a given class can be a good biomarker of some PUFA, but not of others. In particular, a-linolenic acid intake is correlated with its percentage in plasma cholesteryl esters (Ma et al, 1995b;Katan et al, 1997;Amiano et al, 2001;Iwamoto et al, 2002), but not with the small amount present in plasma or erythrocyte phospholipids (Ma et al, 1995b;Kobayashi et al, 2001). However, the intake of long-chain n-3 PUFA is mirrored in all blood lipid fractions, especially plasma cholesteryl esters and plasma or red blood cell phospholipids (Ma et al, 1995b;Katan et al, 1997;Kobayashi et al, 2001).…”

A valid semiquantitative food frequency questionnaire to measure fish consumption

“…We were able to identify five studies in the literature that used multiple linear regression analysis to assess the correlation between fish and seafood and omega-3 PUFA's (Bonaa et al, 1992;Bjerve et al, 1993;Svensson et al, 1993;Hjartaker et al, 1997;Amiano et al, 2001). A sixth study used generalized linear modelling (Wakai et al, 2005).…”

“…All used plasma measures of EPA and/or DHA (phospholipids or phosphatidylcholine) as biomarkers. As a measure of fish intake, three studies used frequency of fish meals per week but did not always indicate if this included seafood (Bonaa et al, 1992;Bjerve et al, 1993;Wakai et al, 2005), two used categories of fish and seafood in frequency and quantities per week (Hjartaker et al, 1997;Amiano et al, 2001), and the sixth used quantities of fish consumed per week (Svensson et al, 1993). Four models included combinations of variables to adjust for the effects of age, sex, BMI, smoking, alcohol consumption and vegetable intake (Bonaa et al, 1992;Bjerve et al, 1993;Amiano et al, 2001;Wakai et al, 2005).…”

“…Various authors have commented that, given the higher content of omega-3 PUFA's in fatty fish, differentiation between lean and fatty fish may be of value (Godley et al, 1996;Woods et al, 2002;Wakai et al, 2005). Although previous validation studies have found significant associations between plasma phospholipid omega-3 PUFAs and fish intake despite reportedly low consumption of fatty fish (Bonaa et al, 1992;Amiano et al, 2001), other studies that have assessed oily and lean fish separately have found that oily fish, but not lean, has a significant correlation with omega-3 PUFAs (Svensson et al, 1993;Hjartaker et al, 1997). Our study goes a step further to reveal a significant negative correlation between lean fish and omega-3.…”