VERY LOW INTAKES OF ?-3 FATTY ACIDS INCORPORATED INTO BOVINE MILK REDUCE PLASMA TRIACYLGLYCEROL AND INCREASE HDL-CHOLESTEROL CONCENTRATIONS IN HEALTHY SUBJECTS

Pharmacological Research, Vol. 41, No. 5, 2000doi:10.1006rphrs.2000.0650, available online at http:rrwww.idealibrary.com on

VERY LOW INTAKES OF N-3 FATTY ACIDS INCORPORATED INTOBOVINE MILK REDUCE PLASMA TRIACYLGLYCEROL AND INCREASE

HDL-CHOLESTEROL CONCENTRATIONS IN HEALTHY SUBJECTS

U `FRANCESCO VISIOLI , PATRIZIA RISE, ELENA PLASMATI, FRANCOPAZZUCCONI, CESARE R. SIRTORI and CLAUDIO GALLI

Institute of Pharmacological Sciences, Uni¨ersity of Milan, Via Balzaretti 9, 20133 Milano,Italy

Eight normolipidaemic volunteers, habitual partial skim milk drinkers and non-eaters offish during the study, were given 500 ml dayy1 of partial skim milk for 1 month; they werethen switched to 500 ml dayy1 of a novel commercially available milk preparation,supplying 400 mg of n-3 fatty acids}of which 300 mg were EPAqDHA}and 15 mgvitamin E, for 6 weeks. No changes in plasma lipid parameters were observed after the first

Žrun-in month; at 3 and 6 weeks on the n-3-rich milk, marked increments of plasma EPA 44. Ž .and 31%, respectively and DHA 13 and 31%, respectively were observed. Triacylglycerol

Ž . Ž .TG concentrations decreased by 19% and high-density lipoprotein HDL concentrationsincreased by 19% at 6 weeks; plasma vitamin E rose by 21% while the susceptibility ofplasma to oxidation was unaffected. Correlations were found between plasma EPA or DHAand TG, cholesterol, and HDL. In conclusion, the intake of a milk preparation providinglow amounts of EPAqDHA to healthy individuals led to marked increases of n-3 fattyacids and vitamin E in plasma and in associated favourable changes in HDL and TG.

Q 2000 Academic Press

KEY WORDS: n-3 fatty acids, cholesterol, atherosclerosis, triacylglycerol, eicosapentaenoic acid, docosahex-aenoic acid, fish oil, marine oil, low-density lipoprotein, high-density lipoprotein.

INTRODUCTION

A number of controlled studies have shown that theŽintake of the n-3 fatty acids eicosapentaenoic EPA,

. Ž .20:5 and docosahexaenoic DHA, 22:6 acids resultsin beneficial effects on various physiological parame-ters, the most relevant being a marked reduction of

Ž .plasma triacylglycerol TG concentrations. Such ef-fects have been generally obtained through pro-longed supplementation of 1 or more g dayy1 ofEPA andror DHA, marketed either as TG or ethyl

Ž .ester EE . It should be noted, however, that retro-spective studies indicate that the consumption ofjust a few fish meals per week can also reduce

w xcardiovascular mortality 1, 2 . The latter observa-tions suggests that the intake of n-3 fatty acids withfish, where they are dispersed in tissue and depotlipids, may be more efficient than when they areadministered as a bolus, i.e. as TG or EE capsules.

Recently, the administration of only 1 g dayy1 of

U Corresponding author.

1043]6618r00r050571]06r$35.00r0

n-3 fatty acids for 3 years led to a highly significantŽ .15% reduction of sudden deaths in otherwise opti-

w xmally treated coronary patients 3 .Lipid-soluble compounds, including fatty acids and

certain drugs, are better absorbed when they areadministered in association with fats and when theyare subdivided into smaller doses throughout the

w xday 4 , circumstances that rarely occur. Milk is themost efficient medium for fat absorption, because ofa high degree of dispersion of the latter in globules,provided in a large volume. This is physiologicallyvery relevant as, for instance, the human newbornrequires a very high fat intake to sustain the rapidpost-natal growth; human milk has the highest en-

Ž y1ergy density approx. 60 kcal g protein, mostly.deriving from fat and provides appreciable amounts

Ž y1of DHA approx. 6]7 mg dl , resulting in an aver-y1 . w xage daily intake of 10 mg kg body wt. 5 .

We studied the effects of a novel bovine milkpreparation enriched in n-3 fatty acids and vitaminE on plasma fatty acids, lipidsrlipoproteins, andvitamin E concentrations, and on the susceptibilityof plasma to oxidation in normal subjects.

Q 2000 Academic Press

Pharmacological Research, Vol. 41, No. 5, 2000572

Table I( ) ( ) ( )Plasma concentrations of total cholesterol TC , triacylglycerol TG , HDL cholesterol HDL-C , and vitamin E, and rate

( )of conjugated diene CD formation

Variable Unit T T T Ty4 0 3 6

y1TC mmol l 4.78"0.22 4.85"0.21 5.22"0.18 4.83"0.18y1 U UTG mmol l 0.65"0.11 0.67"0.14 0.57"0.12 0.55"0.14y1 UUU UUUHDL-C mmol l 1.49"2.14 1.51"2.04 1.80"0.25 1.79"0.24

y1 UU UUvitamin E mg ml 10.1"1.1 10.5"1.2 12.7"1.4 12.6"1.6y1CD Dabs min 16"2 17"2 15"1 18"2

Values are the means"SD. T , 1 month before supplementation with Plus Omega 3W; T , initial values; T , after 3y4 0 3weeks of treatment; T , after 6 weeks of treatment. Values marked are statistically different from those at T as follows:6 0UP-0.05; UUP-0.03; UUUP-0.005.

MATERIALS AND METHODS

Study designŽEight normolipidaemic volunteers 4 M and 4 F,

.27.1"1.7 years gave informed consent to the study.Ž .One month T before being supplemented withy4

Parmalat Plus Omega 3W milk, subjects}who wereregular partially skimmed milk drinkers}were in-structed not to change their habitual diet but toavoid eating fish until the end of the study; theywere supplied with partially skimmed milk, i.e. 1.7 gof fat dly1, of which they were told to drink 500 mldayy1. The subjects at study were all regular milkdrinkers and the consumption of partially skimmedmilk per se in these volunteers had no effect onplasma fatty acids, triacylglycerol, and HDL choles-terol, as shown by the constancy of these parametersduring the run-in month before the supplementation

W Ž .with Plus Omega 3 Table I .At T , subjects were instructed to replace the0

partially skimmed milk with 500 ml dayy1 of thecommercially available Parmalat Plus Omega 3W

milk. This is a partial skim UHT milk preparation,containing 1.7 g dly1 of total fat, including 80 mg

y1 Ž y1dl of n-3 fatty acids of which 20 mg dl were alinolenic acid, 24 mg dly1 EPA and 36 mg dly1

. y1DHA and 3 mg dl of vitamin E. The lipid contentand fatty acid composition was confirmed by GLCŽ .not shown . Therefore, daily intakes of milk fat

Žwere: 8.5 g of total fat, 400 mg of n-3 fatty acids 120.mg as EPA and 180 mg as DHA ; in addition, the

milk provided 15 mg dayy1 of vitamin E.Blood was drawn after an overnight fast 30 days

Ž .T before the supplementation with Parmalaty4W Ž . Ž .Plus Omega 3 and then on day 0 T , and at 3 T0 3

Ž .and 6 T weeks.6Ž .Plasma TG, total cholesterol TC , and HDL-

Ž .cholesterol HDL-C were determined by commer-Žcially available enzymatic kits Boehringer

.Mannheim, Germany . Plasma lipids were extractedŽ .with chloroformrmethanol 2:1 vrv and fatty acid

methyl esters were prepared by acid transmethyla-tion and quantified by GLC employing a capillary

Ž . w xcolumn SP-2330, Supelco, Bellefonte, PA, USA 6 .

Table IIPercentage of the major plasma fatty acids at the three

time points considered in this study

Fatty acid T T T0 3 6

14:0 0.58"0.14 0.50"0.14 0.49"0.1216:0 21.11"1.27 19.94"1.33 20.20"1.3518:0 7.83"0.22 8.35"0.27 8.66"0.2620:0 0.33"0.01 0.36"0.02 0.39"0.0222:0 0.77"0.05 0.85"0.06 0.93"0.0524:0 0.90"0.14 0.81"0.09 1.05"0.1416:1 1.58"0.23 1.36"0.21 1.34"0.1618:1 19.91"0.48 18.94"0.70 18.70"0.3520:1 0.17"0.02 0.15"0.03 0.15"0.0224:1 1.19"0.10 1.43"0.18 1.56"0.1420:3 n-9 0.22"0.03 0.26"0.02 0.40"0.0318:2 n-6 31.07"1.13 32.03"0.99 31.10"0.8018:3 n-6 0.41"0.05 0.43"0.04 0.35"0.0520:3 n-6 1.99"0.16 2.08"0.17 1.93"0.1620:4 n-6 7.72"0.25 7.59"0.24 7.65"0.5322:4 n-6 0.34"0.02 0.39"0.04 0.45"0.0222:5 n-6 0.22"0.01 0.23"0.01 0.21"0.0218:3 n-3 0.30"0.02 0.29"0.02 0.28"0.01

U UU20:5 n-3 0.70"0.06 1.01"0.24 0.92"0.0722:5 n-3 0.41"0.04 0.48"0.03 0.43"0.04

UUU22:6 n-3 2.21"0.26 2.51"0.25 2.93"0.29†

Values are the means"SD. T , initial values; T , after0 33 weeks of treatment; T , after 6 weeks of treatment.6Values marked are statistically different from those at T0as follows: UP s 0.0004; UUP s 0.0103; UUUP s 0.0477;†Ps0.017.

Vitamin E was extracted from plasma and was quan-w x Ž .tified by HPLC 7 . Conjugated diene CD forma-

Žtion in plasma incubated with 2,29-azobis 2-Ž .amidinopropane AAPH, 5 mM at 378 for 30 min

was evaluated as an index of its susceptibility tow xoxidation 8 .

Statistical analysesComparison of values among the different time

points was made by paired Student’s t-test analyses,followed by the Bonferroni correction for multipleanalyses. Correlations between relative levels ofplasma fatty acids were evaluated by the use of thePearson’s coefficient, employing the Systat W soft-ware.

Pharmacological Research, Vol. 41, No. 5, 2000 573

RESULTS

Ž .The average basal, i.e. at T , total cholesterol TC ,y4Ž . Ž .triacylglycerol TG and HDL-cholesterol HDL-C

concentrations, not different between men andwomen, were as follows: TC: 186"7, TG: 62"11,

Ž y1 .HDL-C: 57.8"7.1 mmol , means"SD . Valuesrecorded at T were: TC: 188"8, TG: 60"12,0

Ž y1 .HDL-C: 58.6"7.9 mmol , means"SD , i.e. al-most identical to those obtained at T .y4

Ž .Plasma lipids at the end of the 6 weeks Table IŽ .were characterized by a significant ;19%, P-0.05

Žreduction in TG and by a similar increase ;19%,.P-0.005 in HDL-C, as compared with initial val-

Ž .ues T . Conversely, there was no statistically sig-0nificant difference in total cholesterol levels at thethree time points.

The supplementation of 15 mg dayy1 of vitaminE, on top of that already present in the habitual

Ž .diet, resulted in a significant 21%, P-0.03 eleva-tion of plasma a-tocopherol at 3 weeks; vitamin Econcentration remained significantly elevated up to

Ž .Fig. 1. Correlation between plasma triacylglycerol TG and ew xconcentrations. Fatty acids were quantified by GLC 6 , triacylg

6 weeks. The susceptibility of plasma to AAPH-induced oxidation, evaluated as the rate of produc-tion of conjugated dienes, was unaffected by the

Ž .treatment Table I .Plasma levels of n-3 fatty acids were not signifi-

cantly different between T and T , indicating thaty4 0

the consumption of 500 ml dayy1 of partiallyskimmed milk and abstinence from fish during the 1month run-in period did not affect n-3 plasma con-centrations, likely because prolonged washouts arerequired to cause plasma n-3 fatty acid concentra-

w xtions to decline 6, 9, 10 . In particular, DHA con-centrations at T and T were 2.19"0.22% andy4 0

Ž .2.21" 0.26% means " SD , respectively, whereasthose of EPA were 0.72"0.08% and 0.70"0.06%Ž .means"SD , respectively.

The fatty acid percentage composition of plasmalipids at the various time periods during supplemen-tation with Parmalat Plus Omega 3W is shown inTable II. At all time periods, the major fatty acid

Ž .was linoleic acid 18:2n-6 which accounted for morethan 30% of the total, followed by oleic acid, in the

Ž . Ž .icosapentaenoic EPA and docosahexaenoic DHA acidslycerol by an enzymatic kit.

Pharmacological Research, Vol. 41, No. 5, 2000574

range of 18]20% of the total fatty acids. At T and3

T , the proportion of the major fatty acids was not6

significantly modified, as compared to T , except for0

the increments in EPA and in DHA. The finalincreases in EPA and DHA were 33% and 31%,respectively; the increase in the former was more

Ž .elevated 44% at T .3

In order to assess whether changes in plasmalipidsrlipoproteins were associated with the plasmalevels of EPA and DHA throughout the whole studyperiod, regression analyses between these twoparameters were performed.

A negative correlation between EPA or DHA andŽ .TG levels was found Fig. 1 , but the slope was

steeper when EPA was analysed. A positive correla-tion was also found between HDL-C and DHA, but

Ž .not between HDL-C and EPA Fig. 2 . Finally, al-though no change in total cholesterol concentrationwas noted after the treatment, an overall negativecorrelation between plasma TC and DHA was

Ž .recorded at all times Fig. 3 .

Fig. 2. Correlation between plasma high-density lipoproteŽ .docosahexaenoic DHA acids concentrations. Fatty acids we

precipitation with dextran.

DISCUSSION

The administration to normal subjects of as low as300 mg dayy1 of EPAqDHA}incorporated intomilk}for a short period of time, i.e. 3]6 weeks,

Ž .resulted in significant increments in EPA )40%Ž .and DHA )30% levels in plasma lipids. To our

knowledge, no studies demonstrate increments ofthese fatty acids in plasma associated with such lowdoses of n-3 fatty acids. The substantial increases inplasma vitamin E levels following the administrationof only 15 mg dayy1 of a-tocopherol were alsosomewhat unpredicted, as doses of several hundredmilligrams per day are required in order to induce

w xincrements of 50]100% in plasma 11 . However, itshould be noted that plasma concentrations do notcorrelate directly with the total administered doseand that a-tocopherol levels in circulating cells andtissues are better markers of intake than plasma

w xlevels 12, 13 . The increase in plasma vitamin Elevels associated with the administration of Parmalat

Ž . Ž .in cholesterol HDL-C and eicosapentaenoic EPA andw xre quantified by GLC 6 , HDL-C by an enzymatic kit after

Pharmacological Research, Vol. 41, No. 5, 2000 575

Ž . Ž .Fig. 3. Correlation between plasma total cholesterol TC and docosahexaenoic acid DHA concentrations. DHA wasw xquantified by GLC 6 , total cholesterol by an enzymatic kit.

Plus Omega 3W may have contributed to the preven-tion of the enhanced susceptibility of plasma lipidsto oxidation reported in some, but not all, studies of

w xn-3 supplementation 14]18 .The effects on TG and HDL-C concentrations,

obtained after just 3 weeks of treatment, are note-worthy, as comparable reductions of fasting TG con-centrations have been obtained with much higherdoses of EPA and DHA}from 1 to 7 g dayy1 }for

w xseveral weeks 19 . Significant changes in plasmatriacylglycerol concentrations, and platelets and pro-coagulatory parameters in healthy and hyperlipi-daemic Italian individuals have been previously de-scribed by our research group with the administra-

y1 w xtion of 2 and 3 g day of n-3 fatty acids 9, 10, 20 ; areduction of fasting TG concentration of approxi-mately 20% has been reported with the supplemen-

y1 w xtation of 1 g day EPAqDHA for 16 weeks 21 .In turn, treatments with the usual n-3 preparationsrequire higher doses andror longer duration thanthose employed in the present study in order toeffectively reduce plasma TG concentrations.

n-3 Fatty acids, particularly DHA, have been re-ported to concomitantly increase total cholesterol,LDL, and HDL concentrations when administered

y1 w xin amounts of at least 1 g day 22 . Here, asignificant increment of HDL concentrations withthe very low n-3 intakes provided by the Plus Omega3W milk is reported; furthermore, this increment wasmore strictly associated with DHA than with EPAplasma concentrations; on the other hand, no in-crease in total cholesterol concentration wasobserved at any time point. However, it is note-worthy that, despite the lack of effect of the treat-ment on plasma total cholesterol concentration, anegative correlation between DHA and cholesterolconcentration was found at all times. It should bementioned that marked elevations of total choles-terol have been noted after treatments with high

w xamounts of n-3 fatty acids 19 , although correlationswith plasma EPA and DHA were not reported; inour study we have employed much smaller dosesthat led to a lower, yet highly significant, increase inplasma DHA concentrations. Nevertheless, it can behypothesized that the observed effects are at-tributable to the administered n-3 fatty acids, also inlight of the observed correlations between the con-centrations of n-3 fatty acids and other lipids inplasma.

The high effectiveness of such low-dose treatmentwas indeed unexpected. However, it should be takeninto account that milk fat is dispersed in extremelysmall micelles: the average diameter of the fatglobules in homogenized bovine milk is 1]3 mm;estimations indicate that 1 g of fat in milk is dis-persed in 1010 ] 12 micelles, i.e. a very wide surface ofapproximately 1 m2. These features are likely toimprove the absorption and the bioavailability oflipid-soluble compounds, including n-3 fatty acidsand vitamin E. In addition, as compared with thebolus intake usually employed in therapy, milk is afluid that is slowly and repeatedly ingested duringthe day. This may further facilitate the availability,and potentiate the biological actions, of the n-3 fattyacids, in turn rendering them active even at lowdoses.

ACKNOWLEDGEMENTS

Supported by Parmalat s.p.a. Virginia C. Strandedited the manuscript.

REFERENCES

1. Burr ML, Gilbert JF, Holliday RM, Elwood PC,Fehily AM, Rogers S, Sweetnam PM, Deadman NM.

Pharmacological Research, Vol. 41, No. 5, 2000576

Effects of changes in fat, fish, and fibre intakes ondeath and myocardial reinfarction: diet and reinfarc-

Ž .tion trial DART . Lancet 1989; 2: 757]61.2. Dolecek T. Epidemiological evidence of relationships

between dietary polyunsaturated fatty acids and mor-tality in the multiple risk factor intervention trial.Proc Soc Exp Biol Med 1992; 200: 177]82.

3. Valagussa F, Franzosi MG, Geraci E et al. Dietarysupplementation with n-3 polyunsaturated fatty acidsand vitamin E after myocardial infarction: results ofthe GISSI-Prevenzione trial. Lancet 1999; 354:447]55.

4. Lawson LD, Hughes BG. Absorption of eicosapen-taenoic and docosahexaenoic acid from fish oil tria-cylglycerols or fish oil ethyl esters co-ingested with ahigh-fat meal. Biochem Biophys Res Comm 1988; 156:960]3.

5. Marangoni F, Agostoni C, Lammardo AM, GamboniA, Giovannini M, Galli C, Riva E. Polyunsaturatedfatty acid concentrations in human hindmilk arestable throughout 12-month lactiation and provide asustained intake to the infant during exclusive breast-feeding. An Italian study. Br J Nutr in press.

6. Marangoni F, Angeli MT, Colli S, Eligini, Tremoli E,Sirtori CR, Galli C. Changes of n-3 and n-6 fattyacids in plasma and circulating cells of normal sub-

Ž .jects, after prolonged administration of 20:5 EPAŽ .and 22:6 DHA ethyl esters and prolonged washout.

Biochim Biophys Acta 1993; 1210: 55]62.7. Lehman J, Martin HL. Improved direct determina-

tion of alpha and gamma tocopherols in plasma andplatelets by liquid chromatography, with fluorescencedetection. Clin Chem 1982; 28: 1784]7.

8. Esterbauer H, Gebiki J, Puhl H, Jurgens G. The role¨of lipid peroxidation and antioxidants in oxidativemodification of LDL. Free Rad Biol Med 1992; 13:341]90.

9. Tremoli E, Maderna P, Marangoni F, Colli S, EliginiS, Catalano I, Angeli MT, Pazzucconi F, Gian-franceschi G, Davı G, Stragliotto E, Sirtori CR, Galli`C. Prolonged inhibition of platelet aggregation aftern-3 fatty acid ethyl ester ingestion by healthy volun-teers. Am J Clin Nutr 1995; 61: 1]7.

10. Cerbone AM, Cirillo F, Coppola A, Rise P, Stragliotto´E, Galli C, Giordano M, Tremoli E, Di Minno G.Persistent impairment of platelet aggregation fol-lowing cessation of a short course dietary supplemen-tation of moderate amounts of n-3 fatty acid ethylesters. Thromb Haemost 1999; 82: 128]33.

11. Jha P, Flather M, Lonn E, Farkouh M, Yusuf S. Theantioxidant vitamins and cardiovascular disease: a

critical review of epidemiologic and clinical trial data.Ann Int Med 1995; 123: 860]72.

12. Willet WC, Stampfer MJ, Underwood BA, TaylorJO, Hennekens CH. Vitamins A, E, and carotene:effects of supplementation on their plasma levels.Am J Clin Nutr 1983; 38: 559]66.

13. Perugini C, Bagnati M, Cau C, Bordone R, Zoppis E,Paffoni P, Re R, Albano E, Bellomo G. Distributionof lipid-soluble antioxidants in lipoproteins fromhealthy subjects. II. Effects of in ¨ i o supplementa-tion with alfa-tocopherol. Pharmacol Res 2000; 41: inpress.

14. Hau MF, Smelt AH, Bindels AJ, Sijbrands EJ, Vander Laarse A, Onkenhout W, van Duyvenvoorde W,Princen HM. Effects of fish oil on oxidation resis-tance of VLDL in hypertriglyceridemic patients. Ar-terioscler Thromb Vasc Biol 1996; 16: 1197]202.

15. Harats D, Dabach Y, Hollander G, Ben-Naim M,Stein O, Stein Y. Fish oil ingestion in smokers andnonsmokers enhances peroxidation of plasmalipoproteins. Atherosclerosis 1991; 90: 127]39.

16. Hornstra G, Oostenbuurg GS, Vossen RC. Peroxida-tion of low density lipoproteins and endothelial phos-pholipids: effect of vitamin E and fatty acid composi-tion. World Re¨ Nutr Diet 1994; 75: 149]54.

17. Wander RC, Du SH, Thomas DR. Influence of long-chain polyunsaturated fatty acids on oxidation of lowdensity lipoprotein. Prostagland Leukotr Ess FattyAcids 1998; 59: 143]51.

18. Bonanome A, Biasia F, De Luca M, Munaretto G,Biffanti S, Pradella M, Pagnan A. n-3 fatty acids donot enhance LDL susceptibility to oxidation in hy-pertiacylglycerolemic hemodialyzed subjects. Am JClin Nutr 1996; 63: 261]6.

19. Horrocks LA, Yeo YK. Health benefits of docosa-hexaenoic acid. Pharmacol Res 1999; 40: 211]25.

20. Tremoli E, Eligini S, Colli S, Maderna P, Rise P,´Pazzucconi F, Marangoni F, Sirtori CR, Galli C. n-3fatty acid ethyl ester administration to healthy sub-jects and to hyper triglyceridemic patients reducestissue factor activity by adherent monocytes. Atheros-clerosis Thrombosis 1994; 14: 1600]8.

21. Roche HM, Gibney MJ. Postprandial triacylglyc-erolemia: the effect of a low-fat diet with and with-out fish oil supplementation. Eur J Clin Nutr 1996;50: 617]24.

22. Cobiac L, Clifton PM, Abbey M, Belling GB, NestelPJ. Lipid, lipoprotein, and hemostatic effects of fishvs fish-oil n-3 fatty acids in mildly hyperlipidemicmales. Am J Clin Nutr 1991; 53: 1210]16.