Human Nutrition and Metabolism

A Solid Dietary Fat Containing Fish Oil Redistributes Lipoprotein Subclasses without Increasing Oxidative Stress in Men1,2 Tine Tholstrup,*3 Lars I. Hellgren,† Martin Petersen,* Samar Basu,** Ellen Marie Straarup,† Peter Schnohr,‡ and Brittmarie Sandstro¨m* Centre for Advanced Food Studies, *Department Human Nutrition, The Royal Veterinary and Agricultural University, DK-1958 Frederiksberg, Denmark; †Biocentrum-DTU, Biochemistry and Nutrition, Technical University of Denmark, Lyngby, Denmark; **Section of Geriatrics and Clinical Nutrition Research, Faculty of Medicine, Uppsala University, Uppsala, Sweden; and ‡The Copenhagen City Heart Study, Bispebjerg University Hospital, Copenhagen, Denmark

KEY WORDS:

●

fish oil

●

stearic acid

●

LDL subclasses

Dietary fats can be industrially produced to have specific physiologic properties that may benefit risk markers of coronary heart disease (CHD).4 In the Northwestern part of the world, there is a demand (and need) for a healthy alternative dietary fat (stable toward oxidation) that could be a partial substitute for hard margarine, frying fat, and butter. Fatty acids, which neither increase nor decrease LDL cholesterol, are good candidates to substitute for cholesterol-raising SFA. In addition to being stable to oxidation due to its saturation

●

HDL subclasses

●

oxidative stress

and unlike SFA (except for short-chain SFA), stearic acid does not raise cholesterol (1). Oleic acid is considered beneficial compared with the hypocholesterolemic linoleic acid because it does not decrease HDL cholesterol (2– 4). In addition, oleic acid was shown to result in lower oxidative stress than linoleic acid (5– 8). To obtain a more pronounced effect on blood lipids and lipoproteins, specific bioactive fatty acids could be relevant components of a modified dietary fat. The (n-3) fatty acids were shown to lower plasma triacylglycerol (TAG) (9,10). Fish oil has often, but not always, been shown to increase plasma LDL cholesterol (11–13). The observed increase in LDL cholesterol took place in the larger LDL particles (12), whereas the smaller more atherogenic LDL particles were shown to decrease after the intake of fish oil (14). A less beneficial effect of fish oil supplementation may be that the highly unsaturated fatty acids are prone to fatty acid oxidation (15,16), leading to an increased susceptibility to oxidation and atherogenicity of LDL (12). However, results concerning the effect of fish oil on oxidizability are based mainly on in vitro observations (17,18), which may not accurately represent the in vivo situation. On the basis of existing knowledge concerning the effect of individual fatty acids on blood lipids and lipoproteins including some of our previous results (1,19,20), we produced two

1 Presented in part at the XIIth International Symposium on Atherosclerosis, 25–29 June 2000, Stockholm, Sweden [Tholstrup, T., Sandstro¨m, B., Høy, C. E. & Marckmann, P. (2000) Effect of dietary fatty acids modification on blood lipids, lipoproteins and lipoprotein subclasses in men (abs. TuP25:W12, p. 114)], and at the 17th International Congress of Nutrition, 27–31 August, 2001, Vienna, Austria [Tholstrup, T., Sandstro¨m, B., Petersen, M., Hellgren, L., Straarup, E. M. & Høy, C.-E. (2001) Effect of dietary fatty acid modification on blood lipids, lipoproteins, lipoprotein subclasses and in vivo peroxidation in men. Ann. Nutr. Metab. 45 (suppl. 1): 80 (abs.)]. 2 Supported by the Danish Research Agency through the LMC centre for Advanced Food Studies and Geriatrics Research Foundation (S.B.). 3 To whom correspondence should be addressed. E-mail: [email protected]. 4 Abbreviations used: AA, arachidonic acid; apo, apolipoprotein; CETP, cholesterol ester transfer protein; CHD, coronary heart disease; EIA, enzyme immunoassay; E%, energy percent; F, test fat containing fish oil; IDL, intermediate density lipoprotein; MDA, malondialdehyde; O, test fat containing oleic acid; PGF2␣, prostaglandin F2␣.

0022-3166/04 $8.00 © 2004 American Society for Nutritional Sciences. Manuscript received 13 July 2003. Initial review completed 17 July 2003. Revision accepted 26 January 2004. 1051

Downloaded from jn.nutrition.org by guest on June 6, 2013

ABSTRACT There is a demand and need for healthy solid dietary fats. However, synthetic fats can be tailored to contain specific physiologic properties. Our goal was to design dietary solid test fats that would be both beneficial to the atherogenic lipid profile and stable against lipid peroxidation. Sixteen men (age 35–75 y) substituted 80 g of their normal dietary fat intake with test fat for two periods of 21 d each in a double-blind, randomized, crossover study. Although solid, both test fats were low in cholesterol-raising SFA. Test fat “F” contained 5 g/100 g long chain (n-3) fatty acids matched by oleic acid in test fat “O.” Plasma total triacylglycerol (TAG), VLDL TAG, cholesterol in VLDL, and intermediate density lipoproteins (IDL) were lower (P ⬍ 0.05), whereas apolipoprotein (apo) B of the large LDL-2 (d ⫽ 1031–1042 g/L) subclass, and cholesterol of HDL2b subclass, were higher after intake of F than O fat (P ⬍ 0.05). There was no difference in the effect on in vivo oxidation measured as the ratio of plasma isoprostanes F2 to arachidonic acid and urinary isoprostanes, whereas the vitamin E activity/plasma total lipids ratio was higher after intake of F than O (P ⫽ 0.008). In conclusion, a solid dietary fat containing (n-3) PUFA decreased plasma TAG, VLDL, and IDL cholesterol, and redistributed lipoprotein subclasses in LDL and HDL, with a higher concentration of the larger and less atherogenic subfractions. These changes took place without an increase in oxidative stress as measured by in vivo markers. J. Nutr. 134: 1051–1057, 2004.

THOLSTRUP ET AL.

1052

SUBJECTS AND METHODS Design. We performed a double-blind, randomized, dietary intervention with a crossover design in 2 study periods of 3 wk each, separated by a washout period of 3 wk. Human subjects. The subjects were recruited from the database of the Copenhagen City Heart Study (28). We selected subjects that tended to have an atherogenic lipid profile. The selection criteria were as follows: male, age 30 –75 y; plasma TAG ⬎ 1.3 mmol/L; HDL cholesterol ⱕ 1.2 mmol/L; total cholesterol ⬍ 9 mmol/L; total cholesterol:HDL cholesterol ratio ⬎ 4.5. Subjects (n ⫽ 250) were invited to join the study; of those who responded, 19 qualified for the study. Their ages ranged from 36 to 70 y (mean ⫾ SD, 54.4 ⫾ 10.2 y), their body weights ranged from 64 to 110 kg (mean 83.5 ⫾ 12.5 kg) and BMI ranged from 21.9 to 32.5 kg/m2 (mean 26.2 ⫾ 2.9 kg/m2). Four

TABLE 1 Fatty acid composition of the two test fats Fatty acid1

F fat

O fat g/100 g

14:0 15:0 16:0 16:1 18:0 18:1(n-9) 18:1(n-7) 18:2(n-6) 18:3(n-3) 20:1 20:3(n-6) 20:5(n-3) 22:1 22:5(n-3) 22:6(n-3)

1.3 0.1 9.2 1.9 16.6 30.0 2.1 26.6 1.9 2.9 0.3 2.2 1.4 0.3 3.1

0.0 0.0 9.2 0.3 14.9 42.6 1.9 28.6 1.7 0.3 0.3 0.0 0.0 0.0 0.0

1 Fatty acids are characterized by the number of carbon atoms constituting the structural chain of the fatty acid molecule:the number of double bonds within the chain.

of the subjects were smokers. The subjects had a low-to-moderate level of physical activity and they continued at the same level throughout the study. None of the subjects were diabetic. One of the subjects took antihypertensive medicine. During the study two subjects dropped out due to reasons not related to the study. Diets and test fats. During the study, the subjects underwent 2 intervention periods in which they consumed a diet that substituted the 2 test fats, O or F, for ⬃80 g of their daily fat intake. The O fat was a mixture of mango fat, rapeseed, and safflower oils. This particular mixture resulted in a solid fat in which the proportions of lauric, myristic, and palmitic acids were low, whereas the stearic acid content was high. The F fat had a rather similar composition. However 5% of the oleic acid was substituted by (n-3) fatty acids, i.e., 20:5(n-3) and 22:6(n-3), derived from fish oil. Both test fats were solid at 20°C and had similar melting characteristics. The fatty acid composition of the test fats is given in Table 1. The vitamin E concentration in the test diet did not differ from that in the habitual diet according to the food records. Each day, the subjects were provided with a bun, a piece of cake, a small package of spread, and a ready-made dinner. These foods contained 6.4 MJ/d, 45% of energy (E%) as fat, 47E% as carbohydrate, and 7E% protein, and provided the subjects with 80 g of the specific test fat, which was intended to replace 80 g of their habitual fat intake. Before the study, the subjects’ habitual diets were assessed from 4-d weighed food records; on the basis of this assessment, the subjects were instructed how to substitute the test food for part of their daily food intake during the intervention periods. The subjects’ adherence to the dietary advice was tested by assessing their diets from the 4-d weighed food records after 1 wk during each dietary period. All dietary calculations were done using a national database (Dankost, National Food Agency, Denmark). The calculated nutrient content and fatty acid composition of the habitual and intervention diet per day are presented in Table 2. Blood and urine sampling. After an overnight 12-h fast, venous blood was drawn in the morning before the dietary intervention (baseline, i.e., d 1) and at the end of the study period (d 20 and 22). The blood for lipoprotein analysis was collected in tubes containing EDTA and centrifuged at 3000 ⫻ g for 15 min at 20°C. Plasma (3 mL) was stored at 5°C and ultracentrifugation was started within a maximum of 72 h (in samplings from baseline and d 22). VLDL (d ⬍ 1006 g/L), intermediate-density lipoprotein (IDL) (d ⫽ 1006 – 1019 g/L), LDL (d ⫽ 1019 –1063 g/L), and HDL (d ⫽ 1063–1210 g/L) were separated by ultracentrifugation, and LDL and HDL particles were separated into subfractions as previously described (29). In this study, the method was modified by the use a 50.4 Ti rotor with 4-mL open top tubes. Due to these modifications, the density intervals of the LDL subfractions were slightly changed. The density intervals of the LDL apolipoprotein B subfractions were: LDL-1 apolipoprotein (apo)-B, 1019 –1031 g/L; LDL-2 apo-B, 1031–1034 g/L; LDL-3 apo-B, 1034 –1037 g/L; LDL-4 apo-B, 1037–1039 g/L; LDL-5 apo-B, 1039 – 1042 g/L; and LDL-6 apo-B, 1042–1063 g/L. The density intervals of the HDL subfractions were: HDL2b, 1063–1100 g/L; HDL2a, 1100 – 1125 g/L; and HDL3, 1125–1210 g/L. The density intervals were determined by precision refractometry of blank gradients. Blood for determination of fatty acid composition was drawn into tubes containing EDTA and centrifuged at 3000 ⫻ g for 15 min at 5°C. The samples were placed on ice and stored at ⫺80°C until analysis. Urine samples, which were collected over a period of 24 h before and after the intervention periods, were weighed and mixed. Aliquots were kept at ⫺80°C until analysis for free 8-iso-PGF2␣. Lipid and apolipoprotein analysis. We determined cholesterol and TAG concentrations in lipoproteins and cholesterol concentration in lipoprotein subfractions in plasma. Cholesterol and TAG were measured by enzymatic kits (MPR and GPO-PAP, respectively) from Boehringer Mannheim GmbH. We determined apo B in VLDL, IDL, and LDL particles and apo-A 1 in HDL particles in plasma. Apolipoprotein concentrations were measured with immunological kits (UNIMATE) from Roche. All analyses were carried out on a Cobas Mira analyzer from Roche. The fatty acid profiles of plasma TAG and cholesterol ester were determined by a method previously described (30). TAG, total phospholipids, and cholesterol esters of plasma were separated by a TLC and methylated by transesterification catalyzed by BF3 in methanol (31). FAME were analyzed by GLC as described

Downloaded from jn.nutrition.org by guest on June 6, 2013

nutritionally modified test fats, i.e., oleic (O) and fish (F) (fatty acid composition in Table 1). To obtain a cholesterol neutral solid fat, stable to oxidation, we used fat from mango kernels, which has a high content of stearic acid and a low content of other SFA. Stearic acid is unique in that, in contrast to other long-chain SFA, it does not increase plasma cholesterol. It is not clear why stearic acid does not increase cholesterol, but it has been suggested that lower absorption (21–23) and partial desaturation of stearic acid to the cholesterol neutral oleic acid may occur (24). To provide linoleic acid, high-oleic sunflower oil was added because of the cholesterol neutral properties of oleic acid; rapeseed oil, which has cholesterol-lowering properties, was used (25,26). To obtain a specific plasma TAG-lowering effect, we substituted ⬃3 g of oleic acid of fat O with (n-3) fatty acids to obtain test fat F. The amount of (n-3) fatty acids was based on the decreasing effect on plasma TAG shown previously by others (12,27). Our aim was to test the effect of the two modified test fats O and F on blood lipids and lipoproteins, including concentration of subclasses and oxidative stress in men. Before this experiment, in a pilot study, we tested the effects of the modified fat containing fish oil (F) against a control test fat with a fatty acid composition similar to the average Danish dietary fatty acid composition in men of the same age (unpublished).

FISH OIL REDISTRIBUTES LIPOPROTEIN SUBCLASSES

1053

TABLE 2 Calculated nutrient content and fatty acid composition of habitual and experimental diets1 Variable

Habitual2

F fat3

O fat3

P*

11728.9 ⫾ 463.3 12.1 ⫾ 0.5 47.7 ⫾ 1.1 35.7 ⫾ 1.2 3.4 ⫾ 0.8 0.32 ⫾ 0.03 112.4 ⫾ 4.2 31.9 ⫾ 1.4 41.1 ⫾ 2.2 29.7 ⫾ 1.3 84.0 ⫾ 3.7 329.4 ⫾ 16.6 8.5 ⫾ 0.6

0.86 0.67 0.78 0.93 0.84 0.71 0.92 0.74 0.01 0.50 0.89 0.91 0.59

Unit/d Energy, kJ Protein, %E Carbohydrate, %E Fat, %E Alcohol, %E Cholesterol, g Fat, g Saturated Monounsaturated Polyunsaturated Protein, g Carbohydrate, g Vitamin E,4 TE

11838.9 ⫾ 463.3 12.4 ⫾ 0.5 47.3 ⫾ 1.1 35.5 ⫾ 1.2 3.7 ⫾ 0.8 0.31 ⫾ 0.03 111.8 ⫾ 4.2 32.6 ⫾ 1.4 34.2 ⫾ 2.2 30.9 ⫾ 1.3 84.7 ⫾ 3.7 332.0 ⫾ 16.6 8.9 ⫾ 0.6

Values are means ⫾ SEM, n ⫽ 17. * Different from O, P ⬍ 0.05. Habitual is the diet before inclusion in the study. F contained fish oil and O, oleic acid. For fatty acid composition see Table 1. TE: ␣-tocopherol equivalents, 1 TE is the activity of 1 mg of ␣-tocopherol.

(32). If present, fatty acids from 6:0 to 22:6(n-3) were identified from retention times of actual standards (Nu-Chek-Prep). Analysis of oxidation markers. Total plasma (free and esterified) F2-isoprostanes in plasma. F2-isoprostanes were quantified using the 8-isoprostane enzyme immunoassay (EIA) kit from Cayman Chemicals. Lipid extraction from plasma and alkaline hydrolysis of glycerolipids to free bound F2-isoprostanes were performed as described by Wood et al. (33), except that [3H]prostaglandin (PG)F2 was used as an internal standard, and 0.005% BHT (Sigma Chemical) and 0.025% triphenylphosphine (Sigma Chemical) were used as antioxidants, as recommended by Morrow et al. (34). Initial studies showed that the C18-cartridge purification of F2-isoprostanes, which is recommended by the producer of the EIA-kit, resulted in a fraction containing substantial amounts of FFA (results not shown). Because the presence of arachidonic acid (AA) in this fraction represented a risk for 8-iso-PGF2␣ formation during the preparation and analysis, a new purification method that separated the FFA from the F2-isoprostanes was developed. The hydrolysate from the alkaline hydrolysis was acidified below pH 4 with formic acid and applied to a C18-cartridge (Bond-Elut Jr, 500 mg, Varian) which had been activated previously with 5 mL methanol and 5 mL dH2O. After sample application, the cartridge was rinsed with 5 mL distilled water, followed by 2 ⫻ 5 mL hexane:ethylacetate (9:1, v:v). A preconditioned Si-cartridge (Bound-Elut Jr, 500 mg, Varian), was piggy-backed to the C18 cartridge, and the F2-isoprostanes were eluted onto the Si-cartridge with 7.5 mL ethylacetate:hexane (1:1) containing 0.005% BHT. All remaining FFA were eluted from the Si-cartridge with 5 mL CHCl3, and 8-iso-PGF2␣ was eluted with 2 mL methanol. The methanol was evaporated under N2, and the sample was reconstituted in an EIA buffer provided with the EIA kit. No fatty acids were detected with capillary GLC analysis (not shown) in the F2-isoprostane fractions isolated with this method. Using liquid scintillation counting, the recovery rate radioactivity of an aliquot was determined. The remaining part of the sample was analyzed with the 8-iso-PGF2␣ EIA kit according to the instructions from the manufacturer. The result from the EIA-kit was corrected for recovery using the rate calculated from the internal standard. The antiserum used in this assay has 100% cross-reactivity with 8-iso-PGF2␣, 0.2% each with PGF2, PGF3, PGE1, and PGE2, and 0.1% with 6-keto-PGF1. The detection limit of this assay is 4 ng/L (33). RIA of free 8-iso-PGF2␣. Urine samples (50 ␮L) were analyzed for free 8-iso-PGF2␣ by a specific and validated RIA as described elsewhere (35). The cross-reactivity of the 8-iso-PGF2␣ antibody with 15-keto-13, 14-dihydro-8-iso-PGF2␣, 8-iso-PGF2␤, PGF2␣, 15-ketoPGF2␣, 15-keto-13,14-dihydro-PGF2␣, TXB2, 11␤-PGF2␣, 9␤-

PGF2␣, and 8-iso-PGF3␣ was 1.7, 9.8, 1.1, 0.01, 0.01, 0.1, 0.03, 1.8, and 0.6%, respectively. The detection limit of the assay was ⬃23 pmol/L. Total excretion of 8-iso-PGF2␣ was calculated on the basis of total urine volume. Vitamin E analyses. Plasma concentration of ␣-tocopherol was analyzed as described previously (36). Quantification was carried out by comparing peak areas to the area of standard curves obtained with ␣-tocopherol (Sigma Chemical). Statistics. Values before the intervention periods were compared by paired t tests to ensure that there were no differences. We employed a 2-factor analysis of covariance; with a subject and a diet factor) using the baseline value as covariate to compare the effect of the dietary intervention. Although the design of the study does not fully allow for comparisons with baseline values (there was no run-in period) we obtained some important information by comparing baseline values with values after intervention using a paired t test. Due to variance homogeneity and normal distribution, data were not transformed. Data are expressed as means ⫾ SEM. A P-value ⬍ 0.05 was considered significant.

RESULTS Body weights of the men after the two interventions were not different. Plasma lipids and lipoproteins. Intake of F compared with O resulted in 27% lower plasma TAG (P ⫽ 0.001), 32% lower VLDL TAG (P ⬍ 0.05), 30% lower VLDL cholesterol, and 16% lower IDL cholesterol (P ⬍ 0.05; Table 3). The cholesterol content of the large HDL2b subfraction (1063–1100 g/L) was 20% higher after consumption of fat F rather than after O (P ⬍ 0.05). There was a 24 and 23% higher concentration of the two larger LDL subclasses, LDL-2 apo-B and LDL-3 apo-B, respectively, after the men consumed F consumption compared with O (P ⬍ 0.05). Thus, there was an overall remodeling in lipoprotein subfractions after F compared with O (Fig. 1). No other differences were observed. Oxidative stress. The plasma concentration of 8-isoPGF2␣ was analyzed as an indicator of in vivo lipid peroxidation after intake of increased amounts of (n-3) PUFA. The fish oil– enriched fat did not affect 8-iso-PGF2␣ concentrations in the participants, compared with O when calculated as a concentration in plasma (P ⫽ 0.372). After intake of O, the concentration of 8-iso-PGF2␣ was 230.3 ⫾ 34.6 pmol/L and

Downloaded from jn.nutrition.org by guest on June 6, 2013

1 2 3 4

10917.8 ⫾ 499.3 13.7 ⫾ 0.5 45.9 ⫾ 1.2 32.1 ⫾ 1.3 7.1 ⫾ 0.9 0.37 ⫾ 0.03 92.5 ⫾ 6.3 35.2 ⫾ 2.5 26.3 ⫾ 2.4 12.0 ⫾ 2.0 87.9 ⫾ 3.9 297.0 ⫾ 17.9 7.8 ⫾ 1.2

THOLSTRUP ET AL.

1054

TABLE 3 Fasting plasma lipids and lipoprotein concentration in men at baseline and after consuming the fish oil and oleic acid diets for 3 wk each in a randomized, controlled, crossover study1 F fat

O fat

5.73 ⫾ 0.12 0.67 ⫾ 0.08 0.33 ⫾ 0.02 3.51 ⫾ 0.14 1.21 ⫾ 0.05 2.99 ⫾ 0.14 1.83 ⫾ 0.22 1.14 ⫾ 0.12

5.21 ⫾ 0.17# 0.32 ⫾ 0.05*# 0.21 ⫾ 0.02*# 3.43 ⫾ 0.17 1.21 ⫾ 0.06 2.97 ⫾ 0.23 1.19 ⫾ 0.10*# 0.73 ⫾ 0.09*#

5.29 ⫾ 0.17# 0.46 ⫾ 0.06# 0.25 ⫾ 0.02# 3.28 ⫾ 0.18# 1.20 ⫾ 0.07 2.85 ⫾ 0.20 1.64 ⫾ 0.17 1.08 ⫾ 0.15

114.6 ⫾ 6.0 58.4 ⫾ 5.0 81.3 ⫾ 7.3 113.9 ⫾ 8.5 128.8 ⫾ 6.5 234.4 ⫾ 20.0

97.4 ⫾ 5.5# 68.4 ⫾ 5.5* 93.8 ⫾ 8.3* 123.1 ⫾ 10.4 129.0 ⫾ 11.2 187.1 ⫾ 20.4#

0.25 ⫾ 0.02 0.31 ⫾ 0.02 0.65 ⫾ 0.02

0.30 ⫾ 0.03* 0.32 ⫾ 0.03 0.59 ⫾ 0.03

94.3 ⫾ 7.5# 55.1 ⫾ 5.9 76.2 ⫾ 7.8 109.4 ⫾ 10.1 133.5 ⫾ 10.4 203.1 ⫾ 25.4 0.25 ⫾ 0.03 0.28 ⫾ 0.03# 0.66 ⫾ 0.03

1 Values are mean ⫾ SEM, n ⫽ 17. * Different from O, P ⬍ 0.05 (two-factor ANCOVA with the baseline value used as covariate). # Different from baseline, P ⬍ 0.05. 2 Baseline represents pooled values from the 2 values measured the day before each intervention.

was 211.3 ⫾ 34.6 pmol/L after intake of F (Fig. 2a). When normalized to plasma AA concentration, the amount of 8-isoPGF2␣/AA was 576.8 ⫾ 79.2 nmol/mol after intake of O, and 666.7 ⫾ 79.2 nmol/mol after intake of F (P ⫽ 0.081). (Fig. 2b). The data are presented this way because 8-iso-PGF2␣ formed from AA, and the increased (n-3) PUFA intake decreased the AA concentration in the plasma (0.43 ⫾ 0.02 mmol/L) compared with O (0.51 ⫾ 0.02 mmol/L) (P ⬍ 0.0001). Thus, it

FIGURE 1 Changes from baseline in plasma concentrations of apo-B in the LDL subclasses 1– 6 in men after consuming the F and O diets for 3 wk each in a randomized, controlled, crossover study. Values are means ⫾ SEM, n ⫽ 17. *F differs from O, P ⬍ 0.05. For the fatty acid composition of the test fats, see Table 1.

would be expected that the rate of 8-iso-PGF2␣ formation would be slower at a constant level of oxidative pressure after intake of (n-3) PUFA. When 8-iso-PGF2␣ was measured in urine samples, there was no difference between the effect of F and O in the men, i.e., 2.64 and 2.68 nmol/L, respectively (P

FIGURE 2 Plasma F2-isoprostane concentrations without (a) and with (b) adjustment for the concentration of AA; urinary 8-iso-PGF2␣ concentrations (c) and those concentrations multiplied by urinary excretion (d) in men after consuming the F and O diets for 3 wk each in a randomized, controlled, crossover study. Values are means ⫾ SEM, n ⫽ 17. For the fatty acid composition of the test fats, see Table 1.

Downloaded from jn.nutrition.org by guest on June 6, 2013

Lipids, mmol/L Total cholesterol VLDL-cholesterol IDL-cholesterol LDL-cholesterol HDL-cholesterol LDL/HDL-cholesterol Total triacylglycerol VLDL-triacylglycerol LDL subclass apo-B, mg/L LDL-1 apo-B LDL-2 apo-B LDL-3 apo-B LDL-4 apo-B LDL-5 apo-B LDL-6 apo-B HDL subclass cholesterol, mmol/L HDL2b cholesterol HDL2a cholesterol HDL3 cholesterol

Baseline2

FISH OIL REDISTRIBUTES LIPOPROTEIN SUBCLASSES

DISCUSSION A main result of this study was the decreasing effect of the F diet on plasma TAG, VLDL, and IDL cholesterol. In addition, we found that F redistributed lipoprotein subclasses with an increase in the cholesterol concentration of the large HDL2b together with a larger and thus more favorable LDL particle size in men. Another important observation was that (n-3) PUFA did not increase oxidative stress measured as an in vivo marker. Our results confirm the substantial TAG-lowering effect of (n-3) PUFA shown by others (9,10). This effect is possibly due to a reduction in TAG production in the liver and a concom-

TABLE 4 Fasting fatty acid composition of plasma cholesterol esters in men at baseline and after consuming the fish oil and oleic acid diets for 3 wk each in a randomized, controlled, crossover study1 Fatty acid

F fat

O fat g/100 g

14:0 15:0 16:0 16:1(n-7) 18:0 18:1(n-9) 18:1(n-7) 18:2(n-6) 18:3(n-6) 18:3(n-3) 20:3(n-6) 20:4(n-6) 20:5(n-3) 22:6(n-3) Others2 ⌺%

0.56 ⫾ 0.10 0.16 ⫾ 0.04* 10.80 ⫾ 0.80* 2.17 ⫾ 0.89 1.11 ⫾ 0.13* 17.49 ⫾ 1.46* 1.29 ⫾ 0.14* 51.09 ⫾ 2.59* 0.54 ⫾ 0.23* 0.51 ⫾ 0.13 0.56 ⫾ 0.19* 5.33 ⫾ 0.96 5.05 ⫾ 1.53* 1.34 ⫾ 0.24* 1.22 99.55 ⫾ 0.24

0.57 ⫾ 0.18 0.14 ⫾ 0.04 9.99 ⫾ 1.03 2.68 ⫾ 1.75 1.05 ⫾ 0.16 19.47 ⫾ 2.00 1.12 ⫾ 0.18 53.45 ⫾ 4.78 1.00 ⫾ 0.47 0.52 ⫾ 0.22 0.81 ⫾ 0.21 5.47 ⫾ 1.38 1.06 ⫾ 0.54 0.63 ⫾ 0.13 1.32 99.52 ⫾ 0.26

1 Values are Mean ⫾ SD, n ⫽ 17. * Different from O, P ⬍ 0.05. 2 Fatty acids ⱕ 0.5% are presented as others. Fatty acids from C6

to C24 were identified.

itant suppression of VLDL secretion after fish oil treatment (37–39). As a consequence, VLDL and IDL cholesterol, which are strongly associated with concentrations of plasma TAG (40), were lower after F than O. The difference in the LDL subclass pattern, with a higher contribution of the two larger LDL subclasses (LDL subclasses 2–3), agrees with the results of others (12,14). The visual impression of LDL subclasses (e.g., the concentration of apo-B of the different LDL subclasses seen in Fig. 1) after O was a uniform decrease in the concentration of the 6 subclasses, agreeing with the reported LDL cholesterol decrease after O compared with baseline values. Interestingly, F, which did not decrease total LDL cholesterol compared with baseline values, resulted in a conversion of the small dense LDL to large LDL (LDL subclasses 2–3) particles (12,14). Although a causal relation between the LDL subfraction pattern (e.g., particle size) and CHD has not yet been demonstrated, there is increasing evidence that LDL particle size and density are risk factors for CHD (41). The relation between LDL size and risk, which was shown in observational studies (42– 44), may be ascribed to the association between LDL and increased plasma TAG concentration, low HDL cholesterol, and the presence of insulin resistance syndrome (41). Plausible mechanisms for increased atherogenicity are that small dense LDL particles have a decreased affinity for the LDL receptor (42) and are more susceptible to oxidation (45). The beneficial effect of fish oil on LDL subfractions seemed to be related to the downregulation of plasma TAG, probably associated with the transfer from the TAG rich lipoproteins to LDL by cholesteryl ester transfer protein (CETP) (46). Thus, it was shown that the concentration of small dense LDL particles is increased by a plasma TAG ⬎ 1.5 mmol/L (47). Increased plasma TAG will result in increased lipid transfer to LDL and vice versa. The TAG-enriched LDL formed during lipemia are hydrolyzed to small dense LDL by hepatic lipase (48). It seems plausible that the opposite situation with a decrease in plasma TAG, as demonstrated in our study, would result in lower CETP activity, thereby resulting in the formation of larger, less atherogenic, LDL particles, as found by others (49). We observed a higher concentration of cholesterol in the large HDL2b subclass in the men after intake of F than after O, which is in agreement with the findings of others (14). In examining the relation between HDL subclasses and CHD, epidemiologic studies have shown a stronger inverse relation between HDL2 cholesterol than between HDL3 cholesterol and CHD (50,51), disagreeing with results from a single study (52). A decrease in plasma TAG may change HDL subclass composition (53) and hence HDL particle size. However, it is not clear whether TAG enrichment may modulate or regulate the production of the HDL-associated lipoprotein apo-A I, and factors that affect HDL subfractions remain to be identified and described (54). Isoprostanes are prostaglandin-like stable end products that are formed during the peroxidation of PUFAs in the human body (55). 8-Iso-PGF2␣ is a major F2-isoprostane that is specifically formed from peroxidation of AA, and it has been shown that the concentration of F2-isoprostanes correlates with the degree of lipid peroxidation in humans (56). Due to their stability and specificity, they are considered to be one of the most valuable available biomarkers of in vivo lipid peroxidation (57). In the present study, we showed that an increased intake of (n-3) PUFA– enriched dietary F did not alter the 8-iso-PGF2␣ concentration significantly in plasma or urine compared with O. The fact that increased (n-3) PUFA intake does not increase F2-isoprostanes level agrees with

Downloaded from jn.nutrition.org by guest on June 6, 2013

⫽ 0.801) (Fig. 2c). The excretion rates did not differ after the men consumed F (1575.7 ⫾ 213.1 mL/d) and O (1734.5 ⫾ 214.3 mL/d) (P ⫽ 0.379). After adjustment for excretion (24 h), consumption of F resulted in urinary 8-iso-PGF2␣ that tended to be lower after consumption of F than O (P ⫽ 0.096) (Fig. 2d). Plasma vitamin E. Plasma ␣-tocopherol concentration did not differ after consumption of F (1.65 ⫾ 0.07 ␮mol/L) and O (1.58 ⫾ 0.07 ␮mol/L) (P ⫽ 0.522). However, when normalized to total plasma lipid concentrations expressed as ␮mol/mol lipids (estimated as plasma total cholesterol ⫹ plasma TAG), consumption of F (0.26 ⫾ 0.01 ␮mol/mol lipid) resulted in higher values in the men than consumption of O (0.23 ⫾ 0.01 ␮mol/mol lipid) (P ⫽ 0.008). Fatty acid profiles of cholesterol esters in plasma. The results confirmed compliance of the participants to the study protocol because (n-3) PUFA were incorporated to a higher degree after consumption of the F diet and (n-6) PUFA and 18:1(n-9) after consumption of the O diet. The O diet resulted in a slightly, but significantly lower proportion of 16:0 and 18:0 in cholesterol esters compared with the F diet in the men (Table 4).

1055

1056

THOLSTRUP ET AL.

of the functional characteristics required for fats used in industrial food production. In conclusion, a solid dietary fat containing (n-3) PUFA decreased plasma TAG, VLDL, and IDL cholesterol concentration, and redistributed the lipoprotein subpattern with a larger, and thus more favorable, LDL particle size in men. These changes took place without increasing oxidative stress, as measured by in vivo markers. Thus, by modifying the fatty acid composition, the production of less atherogenic solid dietary fats is possible. ACKNOWLEDGMENTS We thank our technician Vivian Anker for lipid analyses, and Grete Peitersen and Nina Kjeldsen from BioCentrum-DTU, The Technical University of Denmark, for analysis of fatty acids and tocopherols, Eva Sejby for urinary isoprostane analysis, and Berit Hoielt and the other staff of the metabolic kitchen. Research assistant Robin D. K. Christensen is thanked for reviewing the statistics.

LITERATURE CITED 1. Tholstrup, T., Marckmann, P., Jespersen, J. & Sandstro¨ m, B. (1994) Fat high in stearic acid favorably affects blood lipids and factor VII coagulant activity in comparison with fats high in palmitic acid or high in myristic and lauric acids. Am. J. Clin. Nutr. 59: 371–377. 2. Mensink, R. P. & Katan, M. B. (1992) Effect of dietary fatty acids on serum lipids and lipoproteins. A meta-analysis of 27 trials. Arterioscler. Thromb. 12: 911–919. 3. Nydahl, M. C., Gustafsson, I. B. & Vessby, B. (1994) Lipid-lowering diets enriched with monounsaturated or polyunsaturated fatty acids but low in saturated fatty acids have similar effects on serum lipid concentrations in hyperlipidemic patients. Am. J. Clin. Nutr. 59: 115–122. 4. Gill, J.M.R., Brown, J. C., Caslake, M. J., Wright, D. M., Cooney, J., Dorothy, B., Hughes, D. A., Stanley, J. C. & Packard, C. J. (2003) Effects of dietary monounsaturated fatty acids on lipoprotein concentrations, compositions, and subfraction distributions and on VLDL apolipoprotein B kinetics: dose-dependent effects on LDL. Am. J. Clin. Nutr. 78: 47–56. 5. Reaven, P., Parthasarathy, S., Grasse, B. J., Miller, E., Almazan, F., Mattson, F. H., Khoo, J. C., Steinberg, D. & Witztum, J. L. (1991) Feasibility of using an oleate-rich diet to reduce the susceptibility of low-density lipoprotein to oxidative modification in humans. Am. J. Clin. Nutr. 54: 701–706. 6. Kratz, M., Cullen, P., Kannenberg, F., Kassner, A., Fobker, M., Abuja, P. M., Assmann, G. & Wahrburg, U. (2002) Effects of dietary fatty acids on the composition and oxidizability of low-density lipoprotein. Eur. J. Clin. Nutr. 56: 72– 81. 7. Svegliati, B. S., Amelio, M., Fiorito, A., Gaddi, A., Littarru, G. & Battino, M. (1999) Monounsaturated diet lowers LDL oxidisability in type IIb and type IV dyslipidemia without affecting coenzyme Q10 and vitamin E contents. Biofactors 9: 325–330. 8. Berry, E. M., Eisenberg, S., Haratz, D., Friedlander, Y., Norman, Y., Kaufmann, N. A. & Stein, Y. (1991) Effects of diets rich in monounsaturated fatty acids on plasma lipoproteins—the Jerusalem Nutrition Study: high MUFAs vs high PUFAs. Am. J. Clin. Nutr. 53: 899 –907. 9. Harris, W. S. & Muzio, F. (1993) Fish oil reduces postprandial triglyceride concentrations without accelerating lipid-emulsion removal rates. Am. J. Clin. Nutr. 58: 68 –74. 10. Marckmann, P., Jespersen, J., Leth, T. & Sandstro¨ m, B. (1991) Effect of fish diet versus meat diet on blood lipids, coagulation and fibrinolysis in healthy young men. J. Intern. Med. 229: 317–323. 11. Harris, W. S. (1996) n-3 fatty acids and lipoproteins: comparison of results from human and animal studies. Lipids 31: 243–252. 12. Suzukawa, M., Abbey, M., Howe, P. R. & Nestel, P. J. (1995) Effects of fish oil fatty acids on low density lipoprotein size, oxidizability, and uptake by macrophages. J. Lipid Res. 36: 473– 484. 13. Eritsland, J., Seljeflot, I., Abdelnoor, M., Arnesen, H. & Torjesen, P. A. (1994) Long-term effects of n-3 fatty acids on serum lipids and glycaemic control. Scand. J. Clin. Lab. Investig. 54: 273–280. 14. Baumstark, M. W., Frey, I., Berg, A. & Keul, J. (1992) Influence of n-3 fatty acids from fish oils on concentration of high- and low-density lipoprotein subfractions and their lipid and apolipoprotein composition. Clin. Biochem. 25: 338 –340. 15. Anttolainen, M., Valsta, L. M., Alfthan, G., Kleemola, P., Salminen, I. & Tamminen, M. (1996) Effect of extreme fish consumption on dietary and plasma antioxidant levels and fatty acid composition. Eur. J. Clin. Nutr. 50: 741–746. 16. Puiggros, C., Chacon, P., Armadans, L. I., Clapes, J. & Planas, M. (2002) Effects of oleic-rich and omega-3-rich diets on serum lipid pattern and lipid oxidation in mildly hypercholesterolemic patients. Clin. Nutr. 21: 79 – 87. 17. Meydani, M., Natiello, F., Goldin, B., Free, N., Woods, M., Schaefer, E., Blumberg, J. B. & Gorbach, S. L. (1991) Effect of long-term fish oil supple-

Downloaded from jn.nutrition.org by guest on June 6, 2013

other recent studies. When Higdon and co-workers (58) compared plasma F2-isoprostanes and free malondialdehyde (MDA) concentrations after intake of high-oleate, high-linoleate or high-(n-3) PUFA diets, MDA concentration decreased after increased (n-3) PUFA intake, whereas the F2isoprostane concentration was unaffected. In another recent study, urinary secretion of F2-isoprostanes actually decreased after increased intake of (n-3) PUFAs, either as capsules (59) or as an increased intake of fish (60). It was also shown recently that oxidation of LDL ex vivo is not enhanced after fish oil feeding when highly specific markers such as the formation of phosphatidylcholine hydroperoxides and cholesteryl linoleate hydroxyperoxides were analyzed (61). Thus, the results from these studies, together with the present results, suggest that a moderate enrichment of the diet with (n-3) PUFA does not increase lipid peroxidation in vivo. This contradicts earlier studies in which less specific markers of oxidation were used, e.g., TBARS as a marker of MDA content and simple diene conjugation (17,18,62). However, several authors (57,63) have questioned the validity of these unspecific methods, and results based solely on them should be viewed with caution. In this context, it should also be emphasized that although more unsaturation enhances the tendency of PUFAs to oxidize in vitro, it has not been established that this also is the case in vivo (57). The observed tendency toward lower urinary isoprostanes (P ⫽ 0.096) after F compared with O is possibly due to the well-known competition between (n-3) fatty acids and (n-6) fatty acids on their indigenous conversion as shown in patients with noninsulin-dependent diabetes mellitus (59). We measured the plasma concentration of ␣-tocopherol because a decrease in its concentration may indicate an increase in oxidation due to the consumption of antioxidants. Because the function of vitamin E is to protect plasma lipids from oxidation (64), we normalized the concentration of ␣-tocopherol to total plasma lipid concentration (total cholesterol ⫹ TAG concentration), which resulted in a higher ratio after intake of F than O. Thus, our vitamin E data do not support an increased susceptibility to oxidation after the intake of fish oil. An aim of this study was to produce solid dietary fats that, in contrast to other hard dietary fats, did not raise cholesterol. We did not include a control fat for comparison. Thus, although both test fats significantly lowered total plasma cholesterol and consumption of O lowered plasma LDL cholesterol concentration, the design of the study did not fully allow for comparisons with baseline values. A disadvantage of the study design was that the participants, in addition to substituting part of their daily fat intake by test fats, also increased their energy intake mainly from fat, compared with their habitual diet. However, because the increase of energy did not differ between the two test diets after the interventions and no apparent side effects of the increased fat intake (probably due to the optimized fatty acid composition of the test fats) were observed, it is unlikely that the increase in energy compared with baseline levels could have affected the variables tested. Modification of the fatty acid composition in dietary fat, as shown in this study by F, may be of special relevance in individuals with an atherogenic lipid profile characterized by increased plasma TAG and a large proportion of small LDL particles. To the best of our knowledge, this new approach in the research field of dietary fatty acids and lipid profile has not been reported previously. In addition, although oils have many of the health-promoting characteristics related to cardiovascular risks, they do not, in contrast to solid fats, possess some

FISH OIL REDISTRIBUTES LIPOPROTEIN SUBCLASSES

42. Lamarche, B., Despre´ s, J.-P., Moorjani, S., Cantin, B., Dagenais, G. R. & Lupien, P.-J. (1996) Triglycerides and HDL-cholesterol as risk factors for ischemic heart disease. Results from the Que´ bec cardiovascular study. Atherosclerosis 119: 235–245. 43. Stampfer, M. J., Krauss, R. M., Ma, J., Blanche, P. J., Holl, L. G., Sacks, F. M. & Hennekens, C. H. (1996) A prospective study of triglyceride level, low-density lipoprotein particle diameter, and risk of myocardial infarction. J. Am. Med. Assoc. 276: 882– 888. 44. Gardner, C. D., Fortmann, S. P. & Krauss, R. M. (1996) Association of small low-density lipoprotein particles with the incidence of coronary artery disease in men and women. J. Am. Med. Assoc. 276: 875– 881. 45. Tribble, D. L., Holl, L. G., Wood, P. D. & Krauss, R. M. (1992) Variations in oxidative susceptibility among six low density lipoprotein subfractions of differing density and particle size. Atherosclerosis 93: 189 –199. 46. Abbey, M. & Nestel, P. J. (1994) Plasma cholesteryl ester transfer protein activity is increased when trans-elaidic acid is substituted for cis-oleic acid in the diet. Atherosclerosis 106: 99 –107. 47. Minihane, A. M., Khan, S., Leigh-Firbank, E. C., Talmud, P., Wright, J. W., Murphy, M. C., Griffin, B. A. & Williams, C. M. (2000) ApoE polymorphism and fish oil supplementation in subjects with an atherogenic lipoprotein phenotype. Arterioscler Thromb. Vasc. Biol. 20: 1990 –1997. 48. Taskinen, M. R., Lahdenpera, S. & Syvanne, M. (1996) New insights into lipid metabolism in non-insulin-dependent diabetes mellitus. Ann. Med. 28: 335–340. 49. Griffin, B. A. (1997) Low-density lipoprotein subclasses: mechanisms of formation and modulation. Proc. Nutr. Soc. 56: 693–702. 50. Ballantyne, F. C., Clark, R. S., Simpson, H. S. & Ballantyne, D. (1982) High density and low density lipoprotein subfractions in survivors of myocardial infarction and in control subjects. Metabolism 31: 433– 437. 51. Gofman, J. W., Young, W. & Tandy, R. (1966) Ischemic heart disease, atherosclerosis, and longevity. Circulation 34: 679 – 697. 52. Stampfer, M. J., Sacks, F. M., Salvini, S., Willett, W. C. & Hennekens, C. H. (1991) A prospective study of cholesterol, apolipoproteins, and the risk of myocardial infarction. N. Engl. J. Med. 325: 373–381. 53. Robinson, D., Ferns, G. A. A., Bevan, E. A., Stocks, J., Williams, P. T. & Galton, D. J. (1987) High density lipoprotein subfractions and coronary risk factors in normal men. Arteriosclerosis 7: 341–346. 54. Colvin, P. L. & Parks, J. S. (1999) Metabolism of high density lipoprotein subfractions. Curr. Opin. Lipidol. 10: 309 –314. 55. Lawson, J. A., Rokach, J. & Fitzgerald, G. A. (1999) Isoprostanes: formation, analysis and use as indices of lipid peroxidation in vivo. J. Biol. Chem. 274: 24441–24444. 56. Morrow, J. D. & Roberts, L. J. (1999) Mass spectrometric quantification of F2-isoprostanes in biological fluids and tissues as measure of oxidant stress. Methods Enzymol. 300: 3–12. 57. Halliwell, B. (2000) Lipid peroxidation, antioxidants and cardiovascular disease: how should we move forward? Cardiovasc. Res. 47: 410 – 418. 58. Higdon, J. V., Liu, J., Du, S. H., Morrow, J. D., Ames, B. N. & Wander, R. C. (2000) Supplementation of postmenopausal women with fish oil rich in eicosapentaenoic acid and docosahexaenoic acid is not associated with greater in vivo lipid peroxidation compared with oils rich in oleate and linoleate as assessed by plasma malondialdehyde and F(2)-isoprostanes. Am. J. Clin. Nutr. 72: 714 –722. 59. Mori, T. A., Puddey, I. B., Burke, V., Croft, K. D., Dunstan, D. W., Rivera, J. H. & Beilin, L. J. (2000) Effect of omega 3 fatty acids on oxidative stress in humans: GC-MS measurement of urinary F2-isoprostane excretion. Redox. Rep. 5: 45– 46. 60. Mori, T. A., Dunstan, D. W., Burke, V., Croft, K. D., Rivera, J. H., Beilin, L. J. & Puddey, I. B. (1999) Effect of dietary fish and exercise training on urinary F2-isoprostane excretion in non-insulin-dependent diabetic patients. Metabolism 48: 1402–1408. 61. Higdon, J. V., Du, S. H., Lee, Y. S., Wu, T. & Wander, R. C. (2001) Supplementation of postmenopausal women with fish oil does not increase overall oxidation of LDL ex vivo compared to dietary oils rich in oleate and linoleate. J. Lipid Res. 42: 407– 418. 62. Harats, D., Dabach, Y., Hollander, G., Ben Naim, M., Schwartz, R., Berry, E. M., Stein, O. & Stein, Y. (1991) Fish oil ingestion in smokers and nonsmokers enhances peroxidation of plasma lipoproteins. Atherosclerosis 90: 127–139. 63. Meagher, E. A. & Fitzgerald, G. A. (2000) Indices of lipid peroxidation in vivo: strengths and limitations. Free Radic. Biol. Med. 28: 1745–1750. 64. Burton, G. W., Joyce, A. & Ingold, K. U. (1982) First proof that vitaminE is major lipid-soluble, chain-breaking antioxidant in human-blood plasma. Lancet 2: 327.

Downloaded from jn.nutrition.org by guest on June 6, 2013

mentation on vitamin E status and lipid peroxidation in women. J. Nutr. 121: 484 – 491. 18. Allard, J. P., Kurian, R., Aghdassi, E., Muggli, R. & Royall, D. (1997) Lipid peroxidation during n-3 fatty acid and vitamin E supplementation in humans. Lipids 32: 535–541. 19. Tholstrup, T., Marckmann, P., Jespersen, J., Vessby, B., Jart, A. & Sandstro¨ m, B. (1994) Effect on blood lipids, coagulation and fibrinolysis of a fat high in myristic acid and a fat high in palmitic acid. Am. J. Clin. Nutr. 60: 919 –925. 20. Tholstrup, T., Sandstro¨ m, B., Hermansen, J. E. & Hølmer, G. (1998) Effect of modified dairy fat on postprandial and fasting plasma lipids and lipoproteins in healthy young men. Lipids 33: 11–21. 21. Bracco, U. (1994) Effect of triglyceride structure on fat absorption. Am. J. Clin. Nutr. 60 (suppl.): 1002S–1009S. 22. Tholstrup, T., Andreasen, K. & Sandstro¨ m, B. (1996) Acute effect of high-fat meals rich in either stearic or myristic acid on hemostatic factors in healthy young men. Am. J. Clin. Nutr. 64: 168 –176. 23. Jones, A. E., Stolinski, M., Smith, R. D., Murphy, J. L. & Wootton, S. A. (1999) Effect of fatty acid chain length and saturation on the gastrointestinal handling and metabolic disposal of dietary fatty acids in women. Br. J. Nutr. 81: 37– 43. 24. Bonanome, A., Bennet, M. & Grundy, S. M. (1992) Metabolic effects of dietary stearic acid in mice: Changes in the fatty acid composition of triglycerides and phospholipids in various tissues. Atherosclerosis 94: 119 –127. 25. Keys, A., Anderson, J. T. & Grande, F. (1965) Serum cholesterol response to changes in the diet. II. The effect of cholesterol in the diet. Metabolism 14: 759 –765. 26. Gustafsson, I.-B., Vessby, B. & Nydahl, M. (1992) Effects of lipidlowering diets enriched with monounsaturated and polyunsaturated fatty acids on serum lipoprotein composition in patients with hyperlipoproteinaemia. Atherosclerosis 96: 109 –118. 27. Morrow, J. D., Frei, B., Longmire, A. W., Gaziano, J. M., Lynch, S. M., Shyr, Y., Strauss, W. E., Oates, J. A. & Roberts, L. J. (1995) Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers. Smoking as a cause of oxidative damage. N. Engl. J. Med. 332: 1198 –1203. 28. Schnohr, P., Jensen, J. S., Scharling, H. & Nordestgaard, B. G. (2002) Coronary heart disease risk factors ranked by importance for the individual and community. A 21 year follow-up of 12 000 men and women from The Copenhagen City Heart Study. Eur. Heart J. 23: 620 – 626. 29. Baumstark, M. W., Kreutz, W., Berg, A. & Keul, J. (1991) Symmetry of the surface, and structure of the central core of human LDL particles, analyzed by X-ray small angle scattering. Adv. Exp. Med. Biol. 285: 123–130. 30. Becker, C. C., Lund, P., Holmer, G., Jensen, H. & Sandstrom, B. (1999) Effects of butter oil blends with increased concentrations of stearic, oleic and linolenic acid on blood lipids in young adults. Eur. J. Clin. Nutr. 53: 535–541. 31. Hoy, C. E. & Holmer, G. (1988) Dietary linoleic acid and the fatty acid profiles in rats fed partially hydrogenated marine oils. Lipids 23: 973–980. 32. Straarup, E. M. & Hoy, C. E. (2000) Structured lipids improve fat absorption in normal and malabsorbing rats. J. Nutr. 130: 2802–2808. 33. Wood, L. G., Fitzgerald, D. A., Gibson, P. G., Cooper, D. M. & Garg, M. L. (2000) Lipid peroxidation as determined by plasma isoprostanes is related to disease severity in mild asthma. Lipids 35: 967–974. 34. Morrow, J. D. & Roberts, L. J. (1997) The isoprostanes: unique bioactive products of lipid peroxidation. Prog. Lipid Res. 36: 1–21. 35. Basu, S. (1998) Radioimmunoassay of 8-iso-prostaglandin F2␣: an index for oxidative injury via free radical catalysed lipid peroxidation. Prostaglandins Leukot. Essent. Fatty Acids 58: 319 –325. 36. Sunesen, V. H., Weber, C. & Holmer, G. (2001) Lipophilic antioxidants and polyunsaturated fatty acids in lipoprotein classes: distribution and interaction. Eur. J. Clin. Nutr. 55: 115–123. 37. Nestel, P. J., Connor, W. E., Reardon, M. F., Connor, S., Wong, S. & Boston, R. (1984) Suppression by diets rich in fish oil of very low density lipoprotein production in man. J. Clin. Investig. 74: 82– 89. 38. Nestel, P. J. (1990) New lipoprotein profiles and coronary heart disease. Improving precision of risk. Circulation 82: 649 – 651. 39. Rambjor, G. S., Walen, A. I., Windsor, S. L. & Harris, W. S. (1996) Eicosapentaenoic acid is primarily responsible for hypotriglyceridemic effect of fish oil in humans. Lipids 31 (suppl.): S45–S49. 40. Tholstrup, T., Sandstro¨ m, B., Bysted, A. & Hølmer, G. (2001) Effect of six dietary fatty acids on postprandial lipid profile, plasma free fatty acids, lipoprotein lipase and cholesterol ester transfer activities in healthy young men. Am. J. Clin. Nutr. 73: 198 –208. 41. Austin, M. A. & Selby, J. V. (1995) LDL subclass phenotypes and the risk factors of the insulin resistance syndrome. Int. J. Obes. Relat. Metab. Disord. 19 (suppl. 1): S22–S26.

1057