Effects of Scandinavian Caviar Paste Enriched with a Stable Fish Oil on Plasma Phospholipid Fatty Acids and Lipid Peroxidation

ORIGINAL COMMUNICATION Effects of Scandinavian caviar paste enriched with a stable fish oil on plasma phospholipid fatty acids and lipid peroxidation

K Engstro¨m1, R Wallin1 and T Saldeen*1

1Department of Surgical Sciences, University of Uppsala, Uppsala, Sweden

Objective: To study the possibility of increasing the very long-chain n-3 polyunsaturated fatty acids, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), in humans by means of consumption of a common food product, Scandinavian caviar paste, suitable for strategic enrichment with a high concentration of these fatty acids, and to measure the potential inducement of lipid peroxidation. Design: A randomized double blind repeated measures experiment. Subjects and interventions: In total, 16 healthy, nonsmoking subjects (eight men and eight women, age 42712 y) were included in the study. Eight consumed 25 g ordinary caviar paste daily for 3 weeks, and eight the same amount of caviar paste enriched with a very stable fish oil (7%, wt/wt). Blood lipids, plasma phospholipid fatty acids and lipid peroxidation were measured. Results: a-Linoleic acid was significantly decreased after intake of both ordinary (À8%, Po0.05) and fish oil caviar (À10%, Po0.05), as was the sum of all n-6 fatty acids (À6%, Po0.05 and À8%, Po0.001, respectively). The fatty acids EPA and DHA, as well as the sum of all n-3 fatty acids, increased significantly in both caviar groups but more in the group given fish oil caviar paste (EPA: +51%, Po0.05 and +100%, Po0.001, respectively; DHA: +24%, Po0.01 and +29%, Po0.001, respectively; sum of n-3:+27%, Po0.05 and +40%, Po0.001, respectively). Lipid peroxidation, measured as the thiobarbituric acidFmalondialde- hyde adduct, was increased by 26% (Po0.05) after intake of ordinary caviar paste, but was unchanged after intake of fish oil-enriched caviar paste. Conclusion: Scandinavian caviar paste is a spread naturally enriched with n-3 polyunsaturated fatty acids that can be included in the diet to achieve an increase in these fatty acids. However, changing to caviar paste enriched with stable fish oil will lead to a considerably greater increase in EPA and DHA. Sponsorship: Swedish Medical Research Council; Cardinova AB, Uppsala, Sweden. European Journal of Clinical Nutrition (2003) 57, 1052–1059. doi:10.1038/sj.ejcn.1601641

Keywords: fish oil; n-3 polyunsaturated fatty acids; enrichment; lipid peroxidation

Introduction scientific documentation of the beneficial effects of the Since the discovery of the low incidence of ischemic heart long-chain n-3 polyunsaturated fatty acids found in seafood disease among Eskimos (Bang & Dyerberg, 1980), the has grown rapidly, and today it is generally recommended that the intake of these disease-preventive fatty acids be increased. However, most Western societies consume little fish. Hence, an increase in fish consumption would *Correspondence: T Saldeen, Department of Surgical Sciences, University involve major dietary changes, and has not proved to be an of Uppsala, Dag Hammarskjo¨lds va¨g 17, 752 37 Uppsala, Sweden. effective way to achieve an increase in the intake of these E-mail: [email protected]. Guarantor: T Saldeen. fatty acids (Mantzioris et al, 2000). To take fish oil as a Contributors: RW organized the production and consumer sensory test supplement is not an alternative at a community level. and made the fatty acid analyses. KE made the rest of the analyses, However, for a smaller population group on a therapeutic, statistical analyses and drafted the manuscript. TS provided advice clinical level, it can probably be a suitable way, especially if during all stages of the study. All authors were involved in the design of the study, organizing of the dietary intervention, and KE and TS in larger quantities need to be consumed. Another, more editing the manuscript. feasible alternative is strategic fish oil enrichment of suitable Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1053 food products that are regularly consumed by a large part of were also instructed to refrain from alcohol consumption for the population. 2 days and from vigorous physical activity for 1 day before In two earlier investigations, we studied the effect of fish the blood sampling. No acetylsalicylic acid (aspirin) or other oil introduced into bread, a food consumed by most people nonsteroidal anti-inflammatory drugs were to be taken for at in large amounts, on some risk factors for ischemic heart least 14 days before the sampling. None of the subjects had disease (Saldeen et al, 1998; Liu et al, 2001). In the present consumed fish oil previous to the study. Before blood study, we have focused on another food suitable for sampling, the subjects had fasted for 10 h. The venous blood enrichment, namely Scandinavian caviar paste. samples were always drawn at about 08:00 to minimize the Scandinavian caviar is a cod roe paste, usually smoked and effect of diurnal variation, without stasis and, after 15 min of mixed with vegetable oil and different spices, and has a long rest, with the subject in the supine position. Plasma and tradition as a popular spread on bread for all ages. Caviar serum were kept at À701C prior to the different analyses, paste is a good choice of food for enrichment with fish oil, as with the exception of blood lipids and glucose, which were it already tastes partly of fish and can thus disguise the taste measured directly. The subjects were asked to bring the tubes of the fish oil. Also, the high content of vegetable oil can with left-over caviar to the last blood sampling appointment. partly be replaced by fish oil without affecting the con- Compliance was assessed on the basis of reports from the sistency. This makes it possible to include a relatively large subjects, the amount of caviar left after the treatment period, amount of fish oil. As it is a popular spread that is used by a and by analyses of fatty acid composition in plasma large population, especially in the Scandinavian countries, it phospholipids before and after the intervention period. is suitable for enrichment with very long-chain n-3 polyunsaturated fatty acids. In the present investigation, we have addressed the Fish oil and caviar composition question of how the intake of ordinary caviar paste and of The fish oil (ESKIMO-3 Food, Cardinova AB, Sweden) is a fish-oil-enriched caviar paste affects plasma phospholipid natural and very stable fish oil. It is stabilized with a mixture fatty acids and lipid peroxidation in normal subjects. of natural antioxidants (Pufanoxs) and contains 4.5 lU/g of vitamin E. It contains about 30% saturated fatty acids, 25% monounsaturated fatty acids and 45% polyunsaturated fatty Materials and methods acids in the form of triacylglycerols, as assayed by gas–liquid Subjects and study design chromatography (Table 1). Of the total fatty acids, 38% are of The participants in the study were 16 healthy, nonsmoking the n-3 type, 18% consist of eicosapentaenoic acid (20:5, volunteers (eight men and eight women), with a mean age of EPA) and 12% consist of docosahexaenoic acid (22:6, DHA). 42 y (range 26–65 y). Informed consent was obtained and the The cholesterol concentration is less than 3 mg/g. The study was conducted in accordance with local ethical fish oil is very stable because of supplementation with the standards and the Declaration of Helsinki. The study was potent antioxidant mixture Pufanoxs in a specific way. designed as a double-blind repeated measures experiment. The peroxide value in the oil is o1 meq/kg and the anisidine The participants were randomly assigned in a blinded test to value is o7, and after exposure to air for several months one of two groups, to receive ordinary caviar paste at room temperature the peroxide and anisidine values (Scandinavian caviar, cod roe paste) or fish oil-enriched are still unchanged (Saldeen et al, 1999). The stability of caviar paste during a test period of 3 weeks. The ordinary the oil has also been confirmed by an independent, caviar paste group consisted of eight subjects (three males, commercial lipid analysis laboratory. During a specific five females) with a mean age of 42712 y (range 30–58 y) step, the fish oil is cleared of undesirable compounds, and the fish oil-enriched caviar paste group consisted of such as environmental pollutants. All tested pollutants, such eight subjects (five males, three females) with a mean age of as dioxins, DDT, PCBs (polychlorinated biphenyls), and 39712 y (range 26–65 y). Exclusion criteria were treatment heavy metals (mercury, lead, cadmium), were below the for diseases, regular medication, nutritional supplements, detection levels (Jacobs & Johnston, 1995; Food Safety pregnancy or obesity (body mass index (BMI) 430 kg/m2). Authority of Ireland, 2002). For additional pretreatment values of the two groups, see The ordinary caviar was made of 65% cod roe paste Tables 3 and 4 before intake of the caviar pastes. Blood (smoked) and 35% rapeseed oil (wt/wt), and in the fish oil- samples were taken before and at the end of the test period. enriched caviar 20% (wt/wt) of the rapeseed oil was replaced The intake of caviar was 25 g daily and the participants were by fish oil. The fish oil caviar was produced by a Swedish carefully instructed as to how to apply the correct amount caviar manufacturer (AB Boviks Konservfabrik, Lysekil, divided on two sandwiches. The amount corresponded to Sweden) and packed in tubes identical to those of the two 100 mm strings of caviar from the tube on each slice of ordinary caviar in order to make it impossible for the bread, and is a reasonably normal amount when used as a participants to know which caviar type they were consum- spread. ing. The fatty acid compositions of the rapeseed oil used in The subjects were carefully instructed not to change their this preparation, the fish oil, the ordinary caviar and the fish diets or their lifestyle during the experimental period. They oil-enriched caviar are shown in Table 2. The amounts of EPA

European Journal of Clinical Nutrition Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1054 Table 1 Fatty acid composition of the fish oil (% of total fatty acids) Lipoprotein (a) (Lp (a)) was determined by an enzyme immunoassay, TintElizet Lp(a) (Biopool AB, Umea˚, Sweden), 14:0 8.3 20:2n-6 0.2 according to the manufacturer’s instructions. 15:0 0.5 16:3n-4 1.8 16:0 15.2 18:3n-3 a-linolenic 0.8 Lipoperoxides, generated in vivo as a result of oxygen free- 18:0 2.6 16:4n-3 3.2 radical-induced lipid peroxidation, were hydrolyzed, and the 20:0 0.2 18:4n-3 3.4 products, including malondialdehyde (MDA), were reacted 22:0 0.1 20:4n-6 arachidonic 0.9 with thiobarbituric acid (TBA) (Young & Trimble, 1991). The Total saturated 26.9 20:4n-3 0.7 14:1n-7 0.2 22:4n-3 0.2 formed MDA–(TBA)2 adduct was measured by reversed-phase 16:1n-7 9.1 20:5n-3 eicosapentaenoic 17.4 high-performance liquid chromatography (HPLC). The 18:1n-9 8.8 21:5n-3 0.7 method used is described in detail by Wong et al (1987); 18:1n-7 2.4 22:5n-3 2.1 however, a few modifications were made. To the plasma 20:1n-9 1.1 22:6n-3 docosahexaenoic 11.7 22:1n-11 1.0 Others 6.2 samples (2 ml) was added 20 ml butylated hydroxytoluene in Total monounsaturated 22.6 Sum of n-6 2.3 ethanol (5%, w/v) immediately after blood sampling, in 18:2n-6 linoleic 1.2 Sum of n-3 40.2 order to prevent the amplification of peroxidation during the assay (Chirico, 1994). A TBA reaction was carried out by mixing phosphoric acid (750 ml, 0.15 M), plasma (500 ml) and TBA solution (250 ml, 42 mM). The reaction mixture was Table 2 Fatty acid composition (% of total fatty acids) of rapeseed oil incubated in a boiling water bath for 60 min and then cooled and fish oil (ESKIMO-3), and of ordinary caviar and fish-oil-enriched (7%, in ice. In order to precipitate proteins and neutralize the wt/wt) caviar solution, the samples (0.5 ml) were mixed with 0.5 ml Fish-oil- sodium hydroxide (1 M)–methanol (9:91, v/v) and centri- Rapeseed Fish Ordinary enriched fuged (9000 Â g, 5 min, 41C). Since the MDA–(TBA)2 adduct Fatty acid oil oil caviar caviar dissociates slowly at neutral or alkaline pH, this step was performed immediately prior to the HPLC run. MDA was Sum of saturated 7.0 26.9 8.8 12.7 calibrated with an 8 mM 1,1,3,3-tetraethoxypropan solution 18:1n-9 oleic 58.5 11.2 54.0 45.3 18:2n-6 linoleic 21.5 1.2 20.3 16.2 in 40% ethanol, which was diluted (40% ethanol) to 18:3n-3 a-linolenic 10.5 0.8 9.9 8.2 different, suitable standard concentrations. The Gilson HPLC 20:5n-3 eicosapentaenoic 0 17.4 1.3 5.2 system used a Waters Nova-Paks C18 (4 mm, 3.9 Â 150 mm) 22:6n-3 docosahexaenoic 0 11.7 2.1 4.5 column fitted with a guard column (Waters Nova-Paks C18, Sum of n-6 21.5 2.3 20.3 16.6 m Sum of n-3 10.5 40.0 13.3 19.0 4 m, 3.9 Â 20 mm). The mobile phase consisted of methanol and potassium phosphate buffer (50 mM, pH 6.8) as a mixture (35:65, v/v), and the flow rate was 1 ml/min and and DHA corresponding to 25 g of caviar intake were 0.11 injection volume 20 ml. Detection was made using a UV/ and 0.18 g, respectively, for the ordinary caviar, and 0.45 and visible detector set at 532 nm. The intra-assay precision was 0.39 g, respectively, for the fish-oil-enriched caviar (calcu- determined by analysis of a pooled plasma sample in lated from the free fatty acids with a conversion factor of multiplicate (n ¼ 4), and the interassay precision was deter- 0.95 specified for Swedish caviar paste according to the mined for the same sample analyzed in multiple assays Swedish National Food Administration tables of fatty acid (n ¼ 3). The coefficients of variation were found to be 9 and content in foods (1998) and a fat content of the specific 12%, respectively. The results are expressed as mM MDA caviar paste of 36%). As shown in Table 2, compared to the equivalents. ordinary caviar, the fish-oil-enriched caviar had higher Plasma plasminogen activator inhibitor-1 (PAI-1) was contents of saturated fatty acids (+31%), EPA (+75%) and assayed amidolytically by means of COATESTs PAI (Kabi DHA (+53%), and lower contents of oleic acid (À19%), Diagnostica, Mo¨lndahl, Sweden), using the chromogenic linoleic acid (À25%) and a-linolenic acid (À21%). The sum of substrate S-2403. The color was read spectrophotometrically n-6 fatty acids was 22% lower and the sum of n-3 fatty acids at 405 nm. was 30% higher in the fish oil-enriched caviar paste Plasma lipids were extracted with chloroform/methanol compared to the ordinary caviar. and submitted to thin-layer chromatography, and the Prior to this study, a blinded consumer sensory test was phospholipids were subsequently recovered after evaluation performed to find a suitable mixture of the fish oil and caviar of the plate. After hydrolysis and methylation, the fatty acid (Meilgaard et al, 1991). methyl esters were separated by gas–liquid chromatography as previously described (Boberg et al, 1985).

Analyses of blood samples Statistical methods Blood lipids and glucose were measured immediately after The SPSS 10.1 software package (SPSS Inc., Chicago, IL, USA) blood sampling, using routine enzymatic methods. was used for the statistical analyses. An independent samples

European Journal of Clinical Nutrition Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1055 t-test was employed in comparing pretreatment values The results of the study are summarized in Figure 1, Tables between the study groups. Comparisons were made within 3 and 4. The pretreatment values did not differ significantly and between groups. Student’s t-test for paired observations between the two groups. The changes in plasma phospho- was used to compare values within the same subjects before lipid fatty acids observed after the treatment period showed and after the intervention period, and independent samples the same tendencies with both ordinary and fish-oil- t-test was used for comparison between the ordinary and the enriched caviar paste (Figure 1). Linoleic acid was signifi- fish-oil-enriched caviar paste groups after intake. Since the cantly decreased after intake of both ordinary (À8%, distribution of Lp(a) was skewed and not improved by Po0.05) and fish oil-enriched caviar paste (À10%, Po0.05), logarithmic transformation, the nonparametric Wilcoxon as was also the sum of all n-6 fatty acids (À6%, Po0.05 and signed ranks test was used instead. Means7standard devia- À8%, Po0.001, respectively). Arachidonic acid, as well as a- tions (s.d.) are presented and P-values o0.05 were adopted as linolenic acid (18:3, n-3, not shown in table), remained significant and the significance levels were two-sided. unchanged in both cases. However, the n-3 fatty acids EPA and DHA, and the sum of all n-3 fatty acids, were significantly increased after the period both on ordinary Results and on fish oil-enriched caviar paste (EPA: +51%, Po0.05 The blinded consumer sensory test resulted in an enrich- and +100%, Po0.001, respectively; DHA: +24%, Po0.01 and ment with fish oil of 7% (wt/wt), with the rapeseed oil +29%, Po0.001, respectively; sum of n-3: +27%, Po0.05 and replaced by fish oil. With this degree of enrichment, the +40%, Po0.001, respectively). After intake of caviar paste, participants could not tell the difference between the two the two groups differed significantly in DHA (Po0.05) and in caviar types. the sum of n-3 fatty acids (Po0.05, Table 3).

Figure 1 Fatty acid composition (% of total fatty acids) of plasma phospholipids before intake compared to after intake of ordinary and fish oil- enriched caviar paste (7%, wt/wt; 25 g/day for 3 weeks). Data are expressed as mean7s.e. and n ¼ 8+8. *Po0.05, **Po0.01, ***Po0.001 compared to before treatment.

European Journal of Clinical Nutrition Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1056 Table 3 Fatty acid composition (% of total fatty acids) of plasma phospholipids before intake compared to after intake of ordinary caviar paste and fish oil-enriched caviar paste (7%, wt/wt; 25 g/day for 3 weeks), and a comparison between the groups after intake

Ordinary caviar paste Fish-oil-enriched caviar paste

Fatty acid Before After P-valuea Before After P-valuea P-valueb

Linoleic 18:2n-6 20.871.7 19.171.5 0.026 21.871.7 19.671.6 0.035 0.528 Arachidonic 20:4n-6 7.870.7 7.970.9 0.910 7.971.50 7.971.2 0.922 0.995 Eicosapentaenoic 20:5n-3 1.270.4 1.870.6 0.048 1.270.3 2.470.6 o0.001 0.063 Docosahexaenoic 22:6n-3 3.770.2 4.670.5 0.002 4.370.7 5.670.9 o0.001 0.020 Sum of n-6 31.871.3 29.871.9 0.011 32.672.5 30.171.8 o0.001 0.761 Sum of n-3 6.070.6 7.671.0 0.012 6.670.8 9.371.4 o0.001 0.015

All data are expressed as mean7s.d. n=8+8. aAfter intake compared to before intake. bAfter fish-oil-enriched caviar paste intake compared to after ordinary caviar paste intake.

No significant changes in blood lipids, glucose, Lp(a) or enriched caviar paste (Figure 1 and Table 3) followed the PAI-1 activity were observed except for a small increase in same direction, but the fish oil caviar paste had a more total serum cholesterol (+5%, Po0.05), serum HDL choles- pronounced positive effect. The comparison between the terol (+10%, Po0.05) and serum LDL cholesterol (+10%, two groups supports the trend towards a greater incorpora- Po0.05) after intake of the fish oil-enriched caviar paste tion of n-3 fatty acids into phospholipids after intake of the compared to before intake. The most sensitive marker of fish fish oil-enriched compared to the ordinary caviar paste. The oil consumption, triacylglycerols, was decreased by 13% after fish oil-enriched caviar contained 75% more of EPA and 53% intake of fish oil-enriched caviar paste and was increased by more of DHA compared to the ordinary caviar (Table 2), but 8% after intake of ordinary caviar paste, but not significantly after intake of the fish oil-enriched caviar the plasma in either case (Table 4). phospholipid EPA was 26% and DHA 18% higher compared

Lipid peroxidation, measured as the MDA–(TBA)2 adduct, to after intake of ordinary caviar. The differences in the showed a significant increase of 26% (Po0.05) after intake of content of EPA and DHA in the two caviar types are reflected the ordinary caviar paste, whereas after intake of the fish oil- in comparable equal increases after intake. However, despite enriched caviar paste there was no significant change the higher amount of DHA than EPA in the ordinary caviar, (Table 4). the increase in plasma phospholipid EPA after intake was higher than the increase in DHA. This was also experienced by Mantzioris et al (2000) when examining the effectiveness Discussion of a diet containing foods enriched with n-3 polyunsaturated The changes in plasma phospholipid fatty acids after intake fatty acids. This can possibly be explained by the high of ordinary caviar paste and those after intake of fish oil- content of a-linolenic acid, which increases tissue EPA but

Table 4 Blood lipids, glucose, PAI-1, Lp(a) and MDA concentrations before intake compared to after intake of ordinary caviar paste and fish oil-enriched caviar paste (7%, wt/wt; 25 g/day for 3 weeks)

Ordinary caviar paste Fish-oil-enriched caviar paste

Before After P-valuea Before After P-valuea

Serum triacylglycerols (mmol/l) 1.2570.72 1.3470.75 0.084 1.4770.88 1.3170.61 0.387 Total serum cholesterol (mmol/l) 5.4270.98 5.3871.05 0.795 4.9370.84 5.1870.97 0.010 Serum HDL cholesterol(mmol/l) 1.3470.50 1.3470.57 1.000 1.0570.32 1.1570.37 0.027 Serum LDL cholesterol (mmol/l)b 3.5870.67 3.4970.62 0.504 3.2170.63 3.5470.70 0.024 Plasma glucose (mmol/l) 5.1770.24 5.0970.51 0.543 5.5271.27 5.6471.53 0.433 Plasma PAI-1 activity (AU/ml) 18.6477.36 20.1677.85 0.271 16.3377.95 16.8179.24 0.830 Plasma Lp(a) (mg/l) 177167 1617172 0.112 3177294 3167293 0.961 Plasma MDA equivalents (mM) 0.3470.22 0.4470.31 0.041 0.2770.09 0.2870,11 0.720

All data are expressed as mean7s.d. n=8+8. aAfter intake compared to before intake. HDL, high-density lipoprotein; LDL, low-density lipoprotein. bTotal cholesterol ÀHDL cholesterol À0.45 Â triacylglycerols.

European Journal of Clinical Nutrition Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1057 not DHA (Sanders & Younger, 1981; Mantzioris et al, 1994). polyunsaturated fatty acids (Kris-Etherton et al, 2000; The fish oil-enriched caviar paste contained both less a- Sanders, 2000). There are some studies in humans concern- linolenic acid than the ordinary caviar, and more EPA than ing lipid peroxidation in plasma or LDL after intake of fish DHA (Table 2), but probably a part of the increase in EPA also oils, using both high (Higdon et al, 2000, 2001) and low results from an a-linolenic acid conversion. The conversion doses of fish oil (Higgins et al, 2001). None of these studies of EPA to DHA can also go the other way from DHA to EPA found an increased lipid peroxidation. On the contrary, (Tocher & Dick, 2001) and it might be preferable with a high there are also some studies using modified fish oil, which EPA content in the food since it can probably slow down this report of an increased lipid peroxidation as well as adverse reversible conversion which leads to a decreasing DHA effects (Johansen et al, 1999; Stalenhoef et al, 2000). Haglund content. et al (1991) reported on differences in plasma MDA after The pretreatment values did not differ significantly intake of large doses of fish oils containing different amounts between the two groups (Table 4), although the mean MDA of vitamin E. Stability measurements of the fish oils used (eg and HDL cholesterol levels in the fish oil-enriched caviar peroxide and anisidine values before and after storage in paste group were somewhat lower than in the ordinary standardized settings, Frankel, 1998) are, however, not caviar group, and the opposite was true for Lp(a). This provided in these studies. The correlation between intake difference is probably due to the limited number of of unstable fish oils or enriched products with in vivo lipid participants in each group (n ¼ 8). In this study, we found a peroxidation in humans needs more investigation. small but significant increase in total serum cholesterol, According to the blinded consumer sensory test, the fish serum HDL cholesterol and serum LDL cholesterol after oil-enriched caviar did not taste different from the ordinary intake of the fish oil-enriched caviar paste compared to caviar. The taste of the ordinary caviar was a good disguise before intake (Table 4). This is in contrast to Haglund et al for the fish oil, and a relatively large amount of fish oil could (1990), who reported a significant decrease in cholesterol therefore be included. The amount of caviar paste consumed and triacylglycerol concentrations to occur after 3 months in daily (25 g) was equivalent to a normal amount of spread healthy subjects taking a large dose (14 g) of fish oil daily. In used on two sandwiches, and was chosen to reflect a realistic subjects with hyperlipidemia, Liu et al (2001) observed a consumption. significant increase in HDL cholesterol and decrease in In Sweden, fish oil has recently been introduced into other triacylglycerols after 4 weeks but not 2 weeks, after intake types of food, for example margarine spread (Gaio, Arla, of fish oil-enriched bread containing a small amount of fish Go¨tene, Sweden) and bread (LEVA, Pa˚gens, Malmo¨, Sweden). oil. Two slices of this bread (62 g) would provide 0.24 g of The concentration of MDA in plasma was measured by an EPA+DHA. If the margarine spread (10 g) was used on these isocratic HPLC method, essentially as described previously two slices, it would add 0.06 g of EPA+DHA. If fish oil- (Wong et al, 1987). This method is simple and achieves enriched caviar paste was used (25 g) on top of this, an extra excellent separation of the MDA–(TBA)2 adduct from con- 0.84 g of EPA+DHA would be included, giving a total intake taminants. However, a small amount of free MDA might be of 1.14 g EPA+DHA. The fish oil-enriched caviar would then found in plasma, apart from the MDA generated by account for 74% of all EPA+DHA included in the two decomposition of lipid hydroperoxides during the acid- sandwiches, whereas the margarine spread would only heating stage, as well as the MDA that is formed during account for 5%. For comparison, a meal of cod (100 g) would prostaglandin metabolism, but these contributions are very give 0.2 g of EPA+DHA and a meal of salmon (100 g) would small (Chirico, 1994). The mean concentration and range of give 1.8 g of EPA+DHA (National Food Administration, plasma MDA were similar to those reported from other 1998). In Sweden, as in many Western countries, the intake studies (Wong et al, 1987; Young & Trimble, 1991). Plasma of fish has decreased considerably in the last century. If four MDA increased significantly after intake of ordinary caviar meals of cod and four meals of salmon were consumed in a paste (Table 4), an effect that might have been caused by the month, together with only one fish oil-enriched caviar paste natural high content of n-3 polyunsaturated fatty acids. If sandwich (as described above) daily, the total intake of the caviar, for some reason, was not stable enough, or if lipid EPA+DHA would be 8.0 g (0.27 g/day) from the fish meals peroxidation was initiated during production or handling, and 17.1 g (0.57 g/day) from the caviar sandwich, the latter this might have resulted in an increased in vivo lipid then accounting for 68% of the total intake. Even if a smaller peroxidation after ingestion. We have no data on the number of enriched sandwiches were consumed, the intake stability of the ordinary caviar paste or the fish oil-enriched, of EPA+DHA would still make a big difference to a diet poor except for the fish oil included. The addition of the in seafood. antioxidant mixture Pufanoxs (included in the fish oil used) There are not many recommendations on fatty acid into the enriched caviar paste might have terminated an intakes that specify EPA and DHA, and they also vary ongoing lipid peroxidation process, leading to less in vivo considerably. However, a group of nutrition scientists has lipid peroxidation after ingestion despite the higher content recently provided guidelines for specific recommendations of EPA and DHA. Concerns have been raised about the for adequate intakes (AI) of EPA and DHA (Simopolous et al, oxidative stress that can result from increased intakes of n-3 1999), which has also been used by Kris-Etherton et al (2000)

European Journal of Clinical Nutrition Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1058 as a target for EPA and DHA increase in the US diet. This malondialdehyde in humans supplemented with vitamin E. J. group suggests that intake of EPA and DHA combined should Nutr. 121, 165–169. be 0.65 g/day (on a 2000 kcal diet, or 0.3% of energy). In Haglund O, Wallin R, Luostarinen R & Saldeen T (1990): Effect of a new fluid fish oil concentrate, ESKIMO-3, on triglycerides, addition, they also recommend an intake of a-linolenic acid cholesterol, fibrinogen and blood pressure. J. Intern. Med. 227, of 2.2 g/day and an upper limit of linoleic acid of 6.7 g/day. 347–353. The intake of only one fish oil-enriched caviar paste Higdon JV, Du SH, Lee YS, Wu, T & Wander RC (2001): Supplementa- sandwich daily will be almost sufficient to meet the tion of postmenopausal women with fish oil does not increase overall oxidation of LDL ex vivo compared to dietary oils rich in recommendations concerning EPA and DHA. oleate and linoleate. J. Lipid Res. 42, 407–418. All possible ways of increasing the amount of EPA and Higdon JV, Liu J, Du SH, Morrow JD, Ames BN & Wander RC (2000): DHA in populations characterized by a low fish intake must Supplementation of postmenopausal women with fish oil rich in be encouraged, since evidence from the vast and fast- eicosapentaenoic acid and docosahexaenoic acid is not associated with greater in vivo lipid peroxidation compared with oils rich in growing scientific research indicates that these fatty acids oleate and linoleate as assessed by plasma malondialdehyde and play a crucial role in the prevention or improvement of F(2)-isoprostanes. Am. J. Clin. Nutr. 72, 714–722. many different diseases, including ischemic heart disease. To Higgins S, Carroll YL, McCarthy SN, Corridan BM, Roche HM, offer strategically enriched foods, like this fish oil-enriched Wallace JMW, O’Brian NM & Morrissey PA (2001): Susceptibility of LDL to oxidative modification in healthy volunteers supplemen- caviar paste, alongside ordinary products on the shelves is ted with low doses of n-3 polyunsaturated fatty acids. Br. J. Nutr. one way among others. However, the population would have 85, 23–31. to be made aware of the benefits of actively choosing these Jacobs MN & Johnston PA (1995): Organochlorine pesticides and specific products, and of most probably paying a higher price PCB residues in pharmaceutical and industrial grade fish oil. for them. Greenpeace Research Laboratories, Technical note 05/95 (Fish oil 1-17). In conclusion, Scandinavian caviar paste is a spread that is Johansen O, Brekke M, Seljeflot I, Adelnoor M & Arnesen H (1999): naturally enriched with EPA and DHA, and can be included N-3 fatty acids do not prevent restenosis after coronary angio- in the diet to achieve a desired increase in these fatty acids. plasty: results from the CART study. Coronary Angioplasty Rest- Furthermore, since caviar paste is a very suitable food for enosis Trial. J. Am. Coll. Cardiol. 33, 1619–1626. Kris-Etherton PM, Shaffer Taylor D, Yu-Poth S, Huth P, Moriarty K, supplementation with a relatively large amount of fish oil, Fishell V, Hargrove RL, Zhao G & Etherton TD (2000): Poly- the use of this strategically enriched caviar can produce a unsaturated fatty acids in the food chain in the United States. Am. considerable increase in the concentrations of EPA and DHA. J. Clin. Nutr. 71, 179S–188S. More research is warranted concerning the oxidative stress Liu M, Wallin R & Saldeen T (2001): Effect of bread containing fish oil on plasma phospholipid fatty acids, triglycerides, HDL that can result from increased intake of n-3 polyunsaturated cholesterol and malondialdehyde in subjects with hyperlipidemia. fatty acids as well as concerning a possible induction of lipid Nutr. Res. 21, 1403–1410. peroxidation during production and storage of enriched Mantzioris E, Cleland LG, Gibson RA, Neumann MA, Demasi M & foods. James MJ (2000): Biochemical effects of a diet containing foods enriched with n-3 fatty acids. Am. J. Clin. Nutr. 72, 42–48. Mantzioris E, James MJ, Gibson RA & Cleland LG (1994): Dietary substitution with an a-linolenic acid-rich vegetable oil increases eicosapentaenoic acid concentrations in tissues. Am. J. Clin. Nutr. Acknowledgements 59, 1304–1309. We are grateful to AB Boviks Konservfabrik, Lysekil, Sweden, Meilgaard MC, Civille B & Carr T (1991): Sensory Evaulation Techniques, 2nd Edition. Boca Raton, FL, USA: CRC Press Inc. for supplying the caviar paste, to Dr I Marklinder for helping National Food Administration (1998): Livsmedelstabell, Fettsyror. to organize the consumer sensory test, and to the subjects for Uppsala, Sweden: Livsmedelsverkets repro. participating in the study. Saldeen T, Engstro¨m K, Jokela R & Wallin R (1999): Importance of in vitro stability for in vivo effects of fish oils. In Natural Antioxidants and Anticarcinogens in Nutrition, Health and Disease, pp 326–330. Cambridge, UK: The Royal Society of Chemistry, Special Publica- References tion 240. Bang HO & Dyerberg J (1980): Lipid metabolism and ischemic heart Saldeen T, Wallin R & Marklinder I (1998): Effects of a small dose of disease in Greenland Eskimos. In Advances in Nutrition Research,ed. stable fish oil substituted for margarine in bread on plasma HH Draper, pp 1–22. New York, USA: Plenum. phospholipid fatty acids and serum triglycerides. Nutr. Res. 18, Boberg M, Vessby B & Croon L-B (1985): Fatty acid composition of 1483–1492. platelets and of plasma lipid esters in relation to platelet function Sanders TAB (2000): Polyunsaturated fatty acids in the food chain in in patients with ischaemic heart disease. Atherosclerosis 58, 49–63. Europe. Am. J. Clin. Nutr. 71, 176S–178S. Chirico S (1994): High-performance liquid chromatography-based Sanders TAB & Younger KM (1981): The effect of dietary supplements thiobarbituric acid tests. Methods Enzymol. 233, 314–318. of o-3 polyunsaturated fatty acids on the fatty acid composition of Food Safety Authority of Ireland (2002): Investigation on PCDDs/ platelets and plasma choline phosphoglycerides. Br. J. Nutr. 45, PCDFs and several PCBs in fish liver oil capsules. Analysis and 613–616. report provided by ERGO Forschungsgesellschaft mbH, Hamburg, Simopolous AP, Leaf A & Salem Jr, N (1999): Essentiality of and Germany. recommended dietary intakes for omega-6 and omega-3 fatty Frankel EN (1998): Lipid Oxidation. West Ferry, Dundee, Scotland: The acids. Ann. Nutr. Metab. 43, 127–130. Oily Press Ltd. Stalenhoef AF, de Graaf J, Wittekoek ME, Bredie SJ, Demacker PN & Haglund O, Luostarinen R, Wallin R, Wibell L & Saldeen T (1991): Kastelein JJ (2000): The effect of concentrated n-3 fatty acids The effects of fish oil on triglycerides, cholesterol, fibrinogen and versus gemfibrozil on plasma lipiproteins, low density lipoprotein

European Journal of Clinical Nutrition Effects of caviar paste enriched with fish oil on lipid peroxidation K Engstro¨m et al 1059 heterogeneity and oxidizability in patients with hypertriglycer- Wong SHY, Knight JA, Hopfer SM, Zaharia O, Leach Jr, CN & idemia. Atherosclerosis 153, 129–138. Sunderman Jr, FW (1987): Lipoperoxides in plasma as measured by Tocher DR & Dick JR (2001): Effects of essential fatty acid deficiency liquid-chromatographic separation of malondialdehyde–thiobar- and supplementation with docosahexaenoic acid (DHA;22:6n-3) bituric acid adduct. Clin. Chem. 33, 214–220. on cellular fatty acid compositions and fatty acyl desaturation Young IS & Trimble ER (1991): Measurement of malondialdehyde in in a cell culture model. Prostaglandins Leukot. Essent. Fatty Acids plasma by high performance liquid chromatography with fluori- 64, 11–22. metric detection. Ann. Clin. Biochem. 28, 504–508.

European Journal of Clinical Nutrition