点击显示 收起
1 From the Research Department of Human Nutrition, Center of Advanced Food Research, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark (TT, MR, and PN); the Faculty of Medicine, Uppsala University, Uppsala, Sweden (SB); the Danish Institute of Agricultural Sciences, Research Centre Foulum, Tjele, Denmark (KS); and the Biochemistry and Nutrition Group, BioCentrum–DTU, Technical University of Denmark, Lyngby, Denmark (EMS)
2 Supported by the Danish Dairy Research Foundation and the Danish Research Development Program for Food Technology. 3 Address reprint requests and correspondence to T Tholstrup, Research Department of Human Nutrition, Royal Veterinary and Agricultural University, 30 Rolighedsvej, Frederiksberg DK-1958, Denmark. E-mail: tth{at}kvl.dk.
ABSTRACT
Background: Evidence suggests that ruminant trans fatty acids (FAs), such as vaccenic acid, do not increase the risk of ischemic heart disease (IHD). However, the effects of ruminant trans FAs on risk markers of IHD have been poorly investigated.
Objective: The objective was to investigate the effect of butter with a naturally high content of vaccenic acid and a concomitantly higher content of monounsaturated FAs on classic and novel risk markers of IHD.
Design: In a double-blind, randomized, 5-wk, parallel intervention study, 42 healthy young men were given 115 g fat/d from test butter that was high in vaccenic acid (3.6 g vaccenic acid/d) or a control butter with a low content of vaccenic acid. Blood and urine samples were collected before and after the intervention.
Results: The intake of the vaccenic acid–rich diet resulted in 6% and 9% lower total cholesterol and plasma HDL-cholesterol concentrations, respectively, than did the intake of the control diet (P = 0.05 and 0.002, respectively), whereas the ratio of total to HDL cholesterol did not differ significantly between the groups. The FA composition of lipid classes reflected the FAs' proportion of the test butter. No other differences were observed.
Conclusions: Butter high in ruminant trans and monounsaturated FAs resulted in significantly lower total and HDL cholesterol than did the control butter with higher amounts of saturated FAs. It may be that the differences were due to the greater content of monounsaturated FAs and the lesser content of saturated FAs in the butter rich in ruminant trans FAs, rather than to the content of vaccenic acid per se.
Key Words: Milk fat • ruminant trans fatty acid • vaccenic acid • men • HDL cholesterol • lipoproteins • isoprostanes • C-reactive protein • insulin
INTRODUCTION
The major trans fatty acid (FA) in milk fat, vaccenic acid (t18:1n–7, 11-trans-octadecenoic acid), makes up only a small amount of the total dietary trans FA intake (1)—1.7% of the total FA content (2). However, the amount of trans FAs in dairy products varies (by weight) from 1% to 6%, depending on the composition of the cow feed (2, 3). The vaccenic acid content also varies by season (3). Unlike industrially produced trans FAs (mainly elaidic acid, 18:1n–9), vaccenic acid has not been shown to be associated with an increased risk of ischemic heart disease (IHD; 4). This finding was contradicted, however, by results from a study in an elderly population (5). Industrially produced trans FAs have been found in some margarines (although currently in negligible amounts), margarine-based products, and oils for frying. These trans FAs are formed when vegetable and marine oils are hardened by partial hydrogenation to adjust firmness and plasticity. Industrially produced trans FAs have been shown to increase serum LDL cholesterol, decrease HDL cholesterol (6-9), and increase IHD risk (10), whereas the effect of vaccenic acid on lipoproteins in humans has not been investigated. Vaccenic acids have been shown to be the major precursor of conjugated linoleic acid (CLA) in ruminant animals (11) and to increase the serum concentration of CLA in humans when given in the diet (12, 13). Thus, the conversion of vaccenic acid to CLA is likely to contribute to the amount of CLA available to the human body, and endogenously formed CLA may exhibit effects similar to those of dietary CLA.
Because it is possible to alter the amount of vaccenic acid in milk fat through cow-feeding procedures (14, 15), we investigated the effects on traditional and novel risk markers of atherosclerosis, oxidative stress, and insulin and glucose of milk fat with a naturally high content of vaccenic acid, given as part of the natural diet. In addition, we wanted to investigate, if possible, the conversion of naturally occurring vaccenic acid from milk fat to CLA in the human body.
SUBJECTS AND METHODS
We conducted a double-blind, randomized, parallel intervention study. The duration of the intervention period was 5 wk, and during this period the Research Department of Human Nutrition provided the participants with foods that differed in FA composition. The participants were stratified according to body mass index (BMI; in kg/m2) into 1 of 2 groups receiving either a diet high in vaccenic acid or a control diet.
Subjects
The baseline characteristics of the 42 men who completed the study are given in Table 1. All men were apparently healthy, as indicated by a medical questionnaire: they were nonsmokers, none had hypertension or a history of atherosclerotic disease, and none were taking any medication. The participants had a low to moderate level of physical activity and maintained the same level throughout the study. They all agreed to refrain from taking any dietary supplements, from donating blood 2 mo before and during the study, and from taking any medication that might interfere with study measurements.
View this table:
TABLE 1. . Baseline characteristics of study participants1
Before the intervention, the participants' habitual diet was assessed by 4-d weighed-food record. The food records were coded, and the energy intake and composition of the diet were calculated by using a national database (Dankost; National Food Agency, Søborg, Denmark). The mean habitual energy intakes were 12.7 MJ/d (range: 8–18 MJ/d); 31% (20–47%) of energy was from fat, 14% (10–23%) was from protein, and 51% (40–64%) was from carbohydrates. There was no significant difference between the 2 groups in habitual dietary intake.
The protocol and aims of the study were fully explained (orally and in writing) to the participants, who gave written informed consent. The Scientific Ethics Committee of the Copenhagen and Frederiksberg approved the research protocol [(KF) 11-138-99).
Diets and test fats
During the intervention period, the participants replaced part of their diet with test foods into which 1 of 2 test fats was incorporated. The test fats were butter produced from milk with different concentrations of vaccenic acid. The FA composition of the milk was changed by adding sunflower seeds to the cow feed. The milk was produced in an experimental herd at the Foulum Research Center of the Danish Institute of Agricultural Sciences. Two types of butter were produced: one type, with a high content of vaccenic acid, was used for the vaccenic acid diet group, and the other type, with a naturally low content of vaccenic acid, was used for the control group. The FA composition of the test fats is shown in Table 2. Each day the participants were provided with 2 bread rolls, a piece of cake, a package of butter, and a cup of chocolate milk. These foods contained 6.9 MJ energy, of which 60% was from fat, 5% from protein, and 35% from carbohydrates. The participants consumed 115 g test fat/d, which replaced most of their habitual fat consumption. The amount of fat given to the participants was equivalent to the average fat intake of Danish men aged 19–34 y (16). On the basis of the first 4-d weighed-food records, the participants were instructed in how to change their diet to consume the test foods without increasing the total fat content of their diet. We tested adherence to the dietary advice by assessing each participant's diet from another 4-d weighed-food record in the second week of the intervention. On the basis of these dietary registrations and body weights, we instructed the participants to change their dietary habits as necessary.
View this table:
TABLE 2. . Fatty acid composition of the test butter rich in vaccenic acid and the control butter
Blood sampling and analysis
After a 12-h overnight fast, venous blood was collected before the intervention period (day 1) and at the end of the study (day 35). Blood for lipoproteins and FA analysis was collected into tubes containing EDTA, which were kept on ice, and the samples were centrifuged at 4°C and 3000 x g for 15 min.
Blood lipids
Plasma for FA analysis was stored at –80°C, and plasma for other analyses was stored at –20°C until the samples were analyzed. LDL and HDL concentrations were assessed by enzymatic colormetric procedure (LDL cholesterol-plus and HDL cholesterol-plus second-generation kits; Roche, Basel, Switzerland) on a Cobas Mira Plus analyzer (Roche Diagnostic, Basel, Switzerland). Cholesterol and triacylglycerol concentrations were measured in plasma by using enzymatic procedures (CHOD-PAP and GPO-PAP, respectively; both kits from Roche) on a Cobas Mira Plus analyzer.
Fatty acid analysis
Total lipids were extracted from the blood samples according to the method of Folch et al (17). Phospholipids and cholesterol esters were separated by thin-layer chromatography, solvent system heptane:isopropanol:acetic acid 95:5:1 (vol:vol). FA methyl esters were prepared by transesterification of boron trifluoride in methanol (14%, vol:vol) according to the method of Morrison and Smith (18).
Analytic gas chromatography analyses were performed by using a HP 6890 gas chromatograph (Hewlett-Packard GmbH, Waldbronn, Germany) equipped with a flame ionization detector. Helium was used as carrier gas at a constant flow rate of 2.0 mL/min, and a split/splitless injector was used with a split ratio of 1:14. The FA methyl esters were analyzed by using a CP Select for FAME capillary column (100 m x 0.25 mm internal diameter; 0.25-µm film thickness; Chrompack, Middleburg, Netherlands). The oven temperature program was as follows: a hold at 50°C for 5 min, followed by an increase by 10°C/min to 165°C and a hold for 40 min, an increase by 1°C/min to 180°C without a hold, and then an increase by 10°C/min to 200°C and a hold for 25 min. The temperature of the injector and the detector was 270°C. Identification and quantification of the FAs were based on standards of FA methyl esters purchased from Nu-Chek Prep Inc (Elysian, MN).
C-reactive protein
Blood for analysis of C-reactive protein (CRP) concentrations was collected into dry tubes; after coagulation, the samples were centrifuged at 3000 x g for 15 min at 20°C. Serum was stored at –20°C until the samples were analyzed. The CRP concentrations were measured by using enhanced turbidimetric immunoassay [CRP-Latex (II) x 2 Seiken; Denka Seiken Co Ltd, Tokyo, Japan) on a Cobas Mira Plus analyzer.
Oxidative stress
Urinary samples were collected over 24-h periods both before and after the intervention period. The volume and density were recorded and the samples were stored at –80°C until they were analyzed for free 8-iso-prostaglandin F2 (PGF2). Urinary samples (50 µL) were analyzed for free 8-iso-PGF2 with the use of a specific and validated radioimmunoassay as described elsewhere (19). The cross-reactivity of the 8-iso-PGF2 antibody with 15-keto-13,14-dihydro-8-iso-PGF2; 8-iso-PGF2ß; PGF2; 15-keto-PGF2, 15-keto-13,14-dihydro-PGF2; thromboxane B2; 11ß-PGF2; 9ß-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. The urinary concentrations of 8-iso-PGF2 were adjusted with total 24-h urine volume.
Hemostatic risk markers
Blood for measurement of factor VII (FVII) coagulant activity (FVII:c) was collected in citrated tubes (kept at room temperature for 1 h) and centrifuged for 20 min at 20°C at 3000 x g. Plasma was pipetted into plastic vials, rapidly frozen, and stored at –80°C. Additional details were described previously (20). Plasma FVII:c was assessed in a one-stage clotting assay. After incubation of 50 µL diluted test plasma (1:10 in tris) and 40 µL human FVII-deficient plasma and human thromboplastin (Thromborel S; Dade Behring, Marburg, Germany), the clotting time was recorded on an ACL-300 automated coagulation analyzer (Instrumentation Laboratory SPA, Milan, Italy), and FVII:c was expressed relative to an activity of 100 by using a 3-point standard curve. Blood for plasminogen activator inhibitor 1 (PAI-1) was collected in cooled tubes containing strong acidic citrate (Stabilyte; Biopool, Umea, Sweden), which were immediately placed on ice, centrifuged at 3000 x g for 15 min at 4°C within 2 h of blood sampling, and then stored at –80°C until they were analyzed. The concentration of PAI-1 (ng/mL) in plasma was analyzed with the use of an enzymatic immunoassay procedure (TintElize PAI-1 kit; Biopool, Umeå, Sweden) with the use of an SLT Rainbow Scanner (SLT–laninstruments GmbH, Grödig, Salzburg, Austria).
Glucose and insulin
Blood for analysis of insulin and glucose concentrations was collected into dry tubes; after coagulation, the samples were centrifuged at 3000 x g for 15 min at 20°C. Serum was stored at –20°C until the samples were analyzed.
Insulin concentrations were measured with the use of a solid-phase, 2-site fluoroimmunometric assay (Auto Delfia Insulin kit B080–101; Wallac, Turku, Finland) and the Auto Delfia system 1235 514. Glucose concentrations were measured with the use of a hexokinase endpoint procedure in serum (Glucoquant Glucose/HK kit; Roche Diagnostics) on a Cobas Mira Plus analyzer.
Statistical analysis
We used a mixed model analysis of covariance to compare the 2 diets. The respective baseline values were used as covariates, and the analyses were thus adjusted for the baseline values of each variable. When necessary, values were log transformed to normalize the distribution of residual and to obtain variance homogeneity. Statistical tests were performed on the transformed data. Transformation was necessary to ascertain the concentrations of c9,t11 CLA in phospholipid and triacylglycerol, the HDL-cholesterol concentrations, the ratio of total to HDL cholesterol, and plasma triacylglycerol, CRP, 8-iso-PGF2, and glucose concentrations. SAS statistical softwar (version 8.2; SAS Institute Inc, Cary, NC) was used for all statistical analyses. All baseline measures are presented as means ± SDs, and all outcome measures are presented as means (adjusted for baseline values) ± SEs.
RESULTS
Dietary intake
No significant changes in body weight were observed during and after the intervention period. Nutrient intakes for the 2 groups were calculated from the 4-d weighed-food records. The results are shown in Table 3. The distribution (% of energy) of protein, carbohydrates, and total fat did not differ between the groups. The vaccenic acid–rich diet group had a higher intake of MUFAs than did the control diet group (P < 0.0001), which corresponded to the difference in FA composition of the test diets.
View this table:
TABLE 3. . Macronutrient intakes after 2 wk of the intervention diet1
Plasma lipids and lipoproteins
Intake of the vaccenic acid diet resulted in total and HDL-cholesterol concentrations that were 6% and and 9% lower (P = 0.05 and 0.002, respectively) than did intake of the control diet (Table 4). Vaccenic acid intake resulted in nonsignificantly (P = 0.14) lower concentrations than did the control diet. There was no significant difference between the effect of the test diets on total:HDL (P = 0.51) or plasma triacylglycerol concentrations (P = 0.30).
View this table:
TABLE 4. . Effect of the 5-wk dietary intervention on fasting blood lipids, lipoproteins, oxidative stress, and inflammatory and hemostatic markers1
Fatty acid profiles of plasma triacylglycerol, cholesterol esters, and phospholipids
FA compositions of the plasma lipid classes triacylglycerol, cholesterol esters (CEs), and phospholipids (Table 5) reflected the FA composition of the test fats, which confirmed that the participants had complied well. Vaccenic acid intake resulted in significantly lower proportions of most long-chain saturated FA (12:0–16:0) in the 3 lipid classes (P < 0.05) and in higher proportions of total trans FAs in triacylglycerol (97%; P < 0.0001) and phospholipid (127%; P < 0.0001) than did intake of the control diet. The number of participants with detectable concentrations of trans FAs in CEs—1 in the control diet group and 5 in the vaccenic acid–rich diet group—was too small to allow a statistical analysis. Vaccenic acid intake resulted in a 110% higher proportion of c9,t11 CLA in triacylglycerol, 170% more in CEs, and 190% more in phospholipids than did the control diet (P < 0.0001). These differences reflected the FA composition of the test butters (Table 2).
View this table:
TABLE 5. . Fasting fatty acid composition of cholesterol esters, phospholipids, and triacylglycerols before and after 5 wk of the experimental diet1
Inflammatory response, oxidative stress, and hemostatic risk markers
There were no differences in the effect of the test diets on serum CRP concentrations (P = 0.67) or urinary excretion of 8-iso-PGF2 (P = 0.93), FVII:c (P = 0.29), or PAI-1 (P = 0.21) (Table 5).
Insulin and glucose concentrations
There was no significant difference between the effects of the test diets on serum insulin (P = 0.14) or serum glucose (P = 0.44) concentrations (Table 5).
DISCUSSION
In the current study, we used cow-feeding strategies to increase the vaccenic acid concentration in milk fat. This resulted in an overall change in FA composition, with an increase in vaccenic acid but also in CLA and oleic and stearic acids and a decrease in palmitic and myristic acids and short-chain FAs. Thus, we compared the effect of milk fat that was high in vaccenic acid, rather than the effect of vaccenic acid per se, with the effect of conventional milk fat with a very low content of vaccenic acid and a different FA composition—ie, a greater amount of saturated fat. The content of vaccenic diet in the vaccenic acid–rich diet was high (3.6 g/d) compared with the estimated average intake in Denmark. Thus, if ruminant trans FAs should have a greater specific effect than did conventional butter, it seems very likely that it would be elucidated in our study.
The intake of the vaccenic acid–rich diet resulted in lower total and HDL-cholesterol concentrations than did intake of the diet with conventional butter. In addition, LDL cholesterol was lower, but not significantly so, in the group consuming the vaccenic acid–rich diet. No other studies have investigated the effect of vaccenic acid on blood lipids in humans. However, one study in rodents reported that increased concentrations of vaccenic acid in the diet had effects on lipoproteins that did not differ significantly from those of a diet rich in oleic acid (18:1, 9c), medium-chain FAs (8:0 and 10:0), or palmitic acid (16:0) (21). The decrease in HDL-cholesterol concentrations could be either an effect of trans FAs, as has been observed for the industrially produced trans FAs that have elaidic acid as the predominant trans FA (6, 22), or a result of the unsaturated FA content of the vaccenic acid–rich diet, because unsaturated FAs result in a lower plasma HDL cholesterol than do SFAs (8, 23). However, the possibility that milk trans FAs could reduce HDL-cholesterol concentrations is considered unlikely in this study, because we observed neither a simultaneous increase in either LDL or total cholesterol, as reported after industrially produced trans FAs (6-8, 24), nor a less favorable ratio of LDL:HDL cholesterol (9, 23). Instead, there was a significant decrease in plasma total cholesterol, a nonsignificantly lower plasma LDL concentration, and no significant difference in plasma total:HDL cholesterol, which is considered a better marker of IHD risk than are the effects on the specific lipoprotein fractions (23). Thus, our results may indicate that ruminant trans FAs seem not to be cholesterolemic, in contrast to industrially produced FAs. However, further research is needed to clarify the role of ruminant trans FAs.
Contrary to most studies that have analyzed the FA composition of plasma triacylglycerol, we included FA incorporation in CEs (when possible) and phospholipids because those values are better markers of dietary changes over weeks (25), whereas FAs in plasma triacylglycerol reflect the FA composition of the food eaten within hours (26). Overall, the FA composition of lipid classes reflected the FA composition of the test diet, as shown previously by others (27, 28) and ourselves (29). Vaccenic acid has been shown to be converted to c9,t11 CLA in animals through a 9-desaturation step (11, 30), and it was shown in humans that vaccenic acid in the diet significantly increases serum c9,t11 CLA concentrations (12, 13). Intake of a diet high in vaccenic acid resulted in a significantly greater increase in the c9,t11 CLA isomer in plasma CEs, phospholipids, and triacylglycerols than did the control diet. Because of technical problems, we could not separate vaccenic acid from other trans FAs, and for that reason we were not able to provide information about the conversion of vaccenic acid to CLA.
With respect to other risk markers included in this study, we focused on the possible effects of vaccenic acid and CLA and not on the overall shift in FA composition from SFAs to MUFAs, unless there was some known indication that a generally higher content of MUFAs than of SFAs may specifically affect the risk marker. We found no differences in CRP concentrations between the 2 diets. The effect of vaccenic acid on CRP has not been studied previously. Although the intake of trans FAs was associated with a significantly higher concentration of CRP in women with higher BMIs than in those with lower BMIs (31), this relation has not been supported by results from clinical trials, in which the effect on CRP concentrations of trans FAs did not differ significantly from that of a diet rich in SFAs (32, 33). The lack of effect of vaccenic acid also agrees with the fact that CLA mixtures containing c9,t11 CLA have not been shown to increase CRP concentrations (34).
In the current study, the difference between the effect of the different diets on oxidative stress, measured as urinary isoprostanes, was not significant. This finding is in disagreement with the increasing effect on oxidative stress observed after intervention with different amounts of vaccenic acid (11-trans 18:1) and 12-trans 18:1, compared with a baseline diet with a high amount of oleic acid (13). We can offer no satisfactory explanation for the difference between the 2 studies. We speculate that the synthetic test fat used in the other study might have possessed greater oxidative properties than did the butter used in our study.
We observed no difference in the effects of the 2 test diets on plasma insulin and glucose concentrations. Thus, neither the overall change from a saturated to a more monounsaturated FA composition nor the amount of trans FAs and CLA after the intake of vaccenic acid fat in this study affected plasma glucose and insulin concentrations. This lack of an effect of trans FAs on glucose and insulin concentrations is in agreement with findings by others (35, 36). In addition, a CLA mix containing c9,t11 CLA has been shown not to alter serum insulin and glucose concentrations in healthy subjects (37, 38).
With regard to the hemostatic risk markers of IHD, there was no difference between the effect of the 2 diets on plasma FVII:c and PAI-1. These findings are in agreement with results from previous studies, which showed no effect of trans FAs or other FAs—such as stearic acid (39) or MUFAs (9)—on hemostatic risk markers, although the effect of c9,t11 CLA on these risk markers has not been extensively examined. One study found no effect of CLA on PAI-1 (38), and, to our knowledge, no studies have examined the effect of CLA on FVII:c.
It would be relevant to compare the effect of vaccenic acid with that of oleic acid, which, unlike SFAs, is "neutral" with respect to important risk markers for IHD. However, this comparison has not yet been possible because of the enormous costs required to produce purified synthetic vaccenic acid products for human intervention studies. An advantage of the current study is that the results, based on cow-feeding procedures, are applicable to a real-life situation. When cows are given feed with a high content of unsaturated FAs—eg, grass, rapeseed, soybean meal, or sunflower—as was the case in the current study, the content of ruminant trans FAs in the milk will increase. At the same time, this type of feed will result in milk fat with a more favorable FA composition—ie, a lower content of cholesterol-raising FA and a higher content of oleic acid. In addition, the content of the bioactive c9,t11 CLA will be increased compared with milk fat low in trans FAs. A disadvantage of the current study was the very high intake of test butter during the intervention. However, the increase in dietary fat during interventions did not differ between the 2 groups, and neither did the body weights after the intervention. In addition, we did not measure concentrations of antioxidants such as vitamin E or other bioactive phytonutrients stemming from the consumption of sunflower oil, and we cannot exclude that these components may have affected the results. If so, dairy fat containing high amounts of vaccenic acid may have a higher content of these beneficial components than does dairy fat low in trans FAs.
In conclusion, butter high in ruminant trans and monounsaturated FAs reduced total and HDL-cholesterol concentrations significantly more than did control butter with higher amounts of saturated fat. We suggest that the differences were due to the greater content of MUFAs and the lesser content of SFAs in the butter rich in ruminant trans FAs, rather than to the vaccenic acid content per se.
ACKNOWLEDGMENTS
We thank our technicians, Ella Jessen and Hanne Lysdal Petersen from the Department of Human Nutrition and Grete Peitersen from the Danish Technical University, for technical assistance. We thank senior scientist Martin Tang Sørensen for his contribution to the production of butter used in this experiment.
TT was the project leader and main contributor to the manuscript; MR and PN were responsible for the acquisition of data; SB performed the oxidative stress analysis; KS was responsible for the production of the test butter; and EMS was responsible for fatty acid analysis. None of the authors had a personal or a financial conflict of interest.
REFERENCES