Influence of three rapeseed oil-rich diets, fortified with a-linolenic acid, eicosapentaenoic acid or

12 Pages·2007·0.19 MB·English

EuropeanJournalofClinicalNutrition(2007)61,314–325
&2007NaturePublishingGroup Allrightsreserved 0954-3007/07 $30.00
www.nature.com/ejcn
ORIGINAL ARTICLE
Influence of three rapeseed oil-rich diets, fortified
with a-linolenic acid, eicosapentaenoic acid or
docosahexaenoic acid on the composition and
oxidizability of low-density lipoproteins: results
of a controlled study in healthy volunteers
S Egert1, V Somoza2, F Kannenberg3,4, M Fobker4, K Krome2, HF Erbersdobler5 and U Wahrburg1
1DepartmentofHumanNutrition,UniversityofAppliedSciences,Muenster,Germany;2GermanResearchCenterofFood
Chemistry,Garching,Germany;3Leibniz-InstituteofArteriosclerosisResearchattheUniversityofMuenster,Muenster,Germany;
4InstituteofClinicalChemistryandLaboratoryMedicine,UniversityofMuenster,Muenster,Germanyand5InstituteofHuman
NutritionandFoodScience,UniversityofKiel,Kiel,Germany
Objective:Tocomparetheindividualeffectsofdietarya-linolenicacid(ALA),eicosapentaenoicacid(EPA)ordocosahexaenoic
acid(DHA) on low-densitylipoprotein (LDL)fatty acidcomposition, exvivo LDLoxidizability and tocopherol requirement.
Design, setting and subjects: A randomized strictly controlled dietary study with three dietary groups and a parallel
design,consistingoftwoconsecutiveperiods.Sixty-onehealthyyoungvolunteers,studentsatanearbycollege,wereincluded.
Forty-eight subjects (13males, 35 females) completed thestudy.
Interventions: Subjects received a 2-week wash-in diet rich in monounsaturated fatty acids (21% energy) followed by
experimental diets enriched with about 1% of energy of ALA, EPA or DHA for 3 weeks. The omega-3 (n-3) fatty acids were
providedwithspecialrapeseedoilsandmargarines.Thewash-indietandtheexperimentaldietswereidentical,apartfromthe
n-3fatty acidcomposition and thetocopherol content, whichwas adjusted tothecontent of dienoic acidequivalents.
Results: Ex vivo oxidative susceptibility of LDL was highest after the DHA diet, indicated by a decrease in lag time ((cid:1)16%,
Po0.001) and an increase in the maximum amount of conjugated dienes (þ7%, Po0.001). The EPA diet decreased the lag
time ((cid:1)16%, Po0.001) and the propagation rate ((cid:1)12%, Po0.01). Tocopherol concentrations in LDL decreased in the ALA
group((cid:1)13.5%,Po0.05)andDHAgroup((cid:1)7.3%,Po0.05).Plasmacontentsoftocopherolequivalentssignificantlydecreased
inallthreeexperimentalgroups(ALAgroup:(cid:1)5.0%,EPAgroup:(cid:1)5.7%,DHAgroup:(cid:1)12.8%).Thecontentofthethreen-3
polyunsaturatedfattyaciddifferentlyincreasedintheLDL:ontheALAdiet,theALAcontentincreasedby89%(Po0.001),on
the EPA diet the EPA content increased by 809% (Po0.001) and on the DHA diet, the DHA content increased by 200%
(Po0.001).Inaddition,theEPAcontentalsoenhanced(withoutdietaryintake)intheALAgroup(þ35%,Po0.01)andinthe
DHA group (þ284%, Po0.001).
Conclusions:DietaryintakeofALA,EPAorDHAledtoasignificantenrichmentoftherespectivefattyacidintheLDLparticles,
withdietaryEPApreferentiallyincorporated.Inthecontextofamonounsaturatedfattyacid-richdiet,ALAenrichmentdidnot
enhanceLDLoxidizability,whereastheeffectsofEPAandDHAonexvivoLDLoxidationwereinconsistent,possiblyinpartdue
to furtherchanges in LDLfattyacid composition.
Correspondence:DrSEgert,DepartmentofHumanNutrition,UniversityofAppliedSciences,Corrensstrasse25,48149Muenster,Germany.
E-mail:[email protected]
Guarantor:SEgert.
Contributors:SEparticipatedindesigningandplanningthestudy,calculatedthediets,headedtheinvestigationduringthedietaryperiod,measuredtheparameters
ofLDLoxidizability,performedthestatisticalanalysisandwrotethepapertogetherwithUW,whoalsocontributedtodesigningandplanningofthestudy.FK
measuredLDLfattyacidcomposition,MFmeasuredLDLtocopherollevelsandKKmeasuredplasmatocopherollevels.UWinitiatedtheprojecttogetherwithVSand
HFE.
Received16November2005;revised8June2006;accepted4July2006;publishedonline13September2006
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
315
Sponsorship:ThisresearchwasfinanciallysupportedbytheGermanFederalMinistryofEducationandResearch(BMBF)within
theproject ‘NAPUS 2000 – healthyfoodstuffs from transgenic rapeseed’.
EuropeanJournalofClinicalNutrition(2007)61,314–325.doi:10.1038/sj.ejcn.1602523;publishedonline13September2006
Keywords: omega-3 fatty acids;rapeseed oil;dietary study; LDLoxidation
Introduction conducted a strictly controlled dietary study in healthy
volunteerswithconventionalfooddietsenrichedwitheither
Oxidative modifications of low-density lipoprotein (LDL) ALA,EPAorDHA.
particles in the arterial wall (LDL oxidation) are believed
to play a crucial role in atherogenesis. The oxidation is
triggered by an attack of free radicals on double bonds
Methods
of unsaturated fatty acids in the LDL, leading to lipid
peroxidation and degradation of the fatty acids (Steinberg
Subjects
etal.,1989;WitztumandSteinberg,2001).TheoxidizedLDL
Seventy-five volunteers, all students living under boarding
particles promote a variety of atherogenic processes, such
school-likeconditions,werescreenedforparticipation.They
as formation of foam cells and endothelial dysfunction
had to meet the following inclusion criteria: a body mass
(Witztum and Steinberg, 2001). It has been shown that the index of less than 28kg/m2, serum total cholesterol values
exvivosusceptibilityofLDLtooxidationislargelydependent
below 7.76mmol/l (o300mg/dl) and serum triglycerides
on dietary factors, namely the content of antioxidants
below 2.26mmol/l (o200mg/dl). Exclusion criteria were
and the fatty acid composition of the diet (Reaven and
smoking,pregnancyandlactation,alcoholabuse,historyof
Witztum,1996).Adietrichinmonounsaturatedfattyacids
atherosclerotic disease, renal, gastrointestinal or endocrine
(MUFAs) leads to an enrichment of MUFAs in the LDL
disturbances.Furtherexclusioncriteriawereallergy,intoler-
particle and thereby to a higher resistance to the oxidative
ance or aversion to foodstuffs contained in the study
processes as compared to diets rich in omega-6 (n-6)
diets.Sixty-onehealthysubjects(17males,44females)were
polyunsaturated fatty acids (PUFAs), especially linoleic
selected for participation. During the study, one subject
acid(C18:2)(Reavenetal.,1993a;Kratzetal.,2002;Nielsen
withdrew because of acute febrile illness. Ten subjects
etal.,2002).
resigned because they were unwilling or unable to comply
Basedontheseresults,itmightbeassumedthattheintake
with the dietary regimen,two students were excluded from
offattyacidswithmorethantwodoublebonds,suchasthe
the study because of non-compliance. The baseline char-
n-3 fatty acids a-linolenic acid (C18:3; ALA), eicosapentae-
acteristics of the 48 subjects (13 males, 35 females, aged
noic acid (C20:5; EPA) and docosahexaenoic acid (C22:6;
between 18 and 45 years) who completed the study are
DHA) would lead to an even higher susceptibility to LDL
presented in Table 1. Only data from these 48 participants
oxidation than linoleic acid. However, studies addressing
were analyzed and included in this report. The subjects
thisquestionhavedemonstratedconflictingresults.Insome
did not take any medications or nutritional supplements
studies, LDLparticles showed anenhanced susceptibility to
(i.e. vitamin E, vitamin C and carotenoids), including fish
oxidation after intake of EPA and DHA (Oostenbrug et al.,
oil before or during the study. Twenty-three women took
1994; Suzukawa et al., 1995; Turini et al., 2001), whereas
oral contraceptives. They were instructed not to stop
other studies revealed no such effect (Frankel et al., 1994;
taking them and not to change to another pill. All subjects
Bonanome et al., 1996; Wander et al., 1996; Higdon et al.,
were requested to maintain their regular lifestyles and
2001;Higginsetal.,2001).Thesecontradictoryfindingsmay
be related to methodical differences in the assays used for
measuring LDL oxidation. Another possible reason for the Table1 Baselinecharacteristicsofthesubjects(n¼48)
contradictoryfindingscouldbethatthemajorityofstudies
Age(years) 25.976.82a
conducted so far used mixtures of EPA and DHA, which Height(m) 1.7270.09a
differedintheirtotalandrelativecontentofEPAandDHA. Weight(kg) 65.779.84a
Furthermore, there are no investigations on the effects of Bodymassindex(kg/m2) 22.272.95a
ALAonexvivoLDLoxidationincomparisontoEPAorDHA. Totalserumcholesterol(mmol/l) 4.9270.71a,b
SerumLDLcholesterol(mmol/l) 2.5470.56a,b
Therefore,themainobjectiveofourstudywastoinvestigate SerumHDLcholesterol(mmol/l) 1.6870.37a,b
the influence of dietary intakes of ALA, EPA and DHA on Serumtriglycerides(mmol/l) 1.14(2.22)c
LDL fatty acid composition and ex vivo oxidizability.
Furthermore, we wanted to evaluate the additional Abbreviations:LDL,low-densitylipoprotein;HDL,high-densitylipoprotein.
aMean7s.d.
tocopherol requirement resulting from the intake of these bToconvertcholesterolvaluestomg/dl,multiplyby38.66.
highly unsaturated fatty acids. For this purpose, we cMedian(range)–toconverttriglyceridestomg/dl,multiplyby88.57.
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
316
especiallytheirusualextentofphysicalactivitiesthroughout dietaryrecordsandthestudydietswerecalculatedusingthe
thestudy. computer-basednutrient-calculationprogramEBISpro(E&D
The protocol and the objectives of the study were Partner,Stuttgart,Germany),basedontheGermanNutrient
explained to the subjects in detail, and all gave their Data Base Bundeslebensmittelschlu¨ssel, version II.3 (BgVV,
written consent. The study protocol was approved by the Berlin, Germany, 1999). The diets were calculated for 10
ethical committee of the University of Muenster and was differentenergylevels,whichrangedfrom7.52to15.05MJ/
in accordance with the Helsinki Declaration of 1975, as day and had a difference of 0.84MJ between each energy
revisedin1983. group. All participants were weighed twice weekly while
wearinglightclothing,andenergyintakewasadjustedwhen
bodyweightchangedmorethan0.5kgtomaintainastable
Designanddiets body weight. During the experimental period, the mean
The study was conducted in a parallel design and consisted body weight not significantly decreased by (cid:1)0.1070.76kg
of a 2-week wash-in period followed by an experimental (mean7s.d.;range(cid:1)1.0to þ1.3kg;P¼0.356).
period of 3 weeks. During the wash-in period, all subjects Thecompositionoftheparticipants’habitualdietandthe
received a diet rich in MUFAs. Thereafter, the subjects were study diets is given in Table 2. The study diets met the
randomlydividedintothreedietgroups:Onegroup(n¼15) German dietary reference values for vitamins and minerals
received a rapeseed oil diet fortified with ALA (ALA (Deutsche Gesellschaft fu¨r Erna¨hrung (DGE) et al., 2000).
group), for the second group (n¼17), the rapeseed oil diet They consisted of conventional mixed foods that were
was fortified with EPA (EPA group) and for the third group freshly prepared. Menus were changed daily. All meals
(n¼16)withDHA(DHAgroup)(seebelow).Therandomiza- were prepared in the kitchen located in the school in
tionwasmadeusingtablesofrandomdigits,generatedbya which the students were trained and housed during the
computer, and separated for men and women. The wash-in week. The subjects ate breakfast and lunch in the dining
diet and the experimental diets were identical in every roomoftheschoolonworkingdays,underthesupervision
aspect, apart from the n-3 fatty acid composition and the of one of the authors (SE). After breakfast, the subjects also
tocopherol content (see below). Venous blood samples gottheirfoodsupplyfortheeveningandbetween-meals.On
wereobtainedatthebeginningofthestudy(visit1),atthe Fridayafternoons,participantswereprovidedwithhampers
end of the wash-in period (visit 2) and after the 3-week containing all their food for the weekend, as well as
experimentalperiod(visit3). detailed written instructions for its preparation. All food-
Before the study, all participants wrote a careful 3-day stuffssuppliedwereweighedforeachindividualtoensurean
dietary record (completed on 2 weekdays and 1 weekend appropriateenergyintakeandaconstantbodyweight.Fatty
day)ofallfoodsandbeveragesconsumed.Thiswasusedto fish such as mackerel, herring and salmon as well as other
estimate each subject’s energy and nutrient intake. The sourcesofEPAandDHA(e.g.n-3fattyacidenrichedfoods)
Table2 Compositionofthehabitualdietandthestudydietsa
Energy/nutrient Wash-inperiod Experimentalperiod
Habitualdiet(n¼48)b Wash-indiet(n¼48) ALAdiet(n¼15) EPAdiet(n¼17) DHAdiet(n¼16)
Totalenergy(MJ/day)c 10.3872.87 10.0371.96 9.3671.95 9.7372.21 10.3871.64
Proteind 13.872.4 14.970.4 14.870.3 14.770.5 14.870.5
Carbohydratesd 46.576.9 44.071.3 44.171.0 44.371.0 44.171.4
Totalfatd,e 37.375.7 40.170.8 40.271.0 40.170.5 40.170.5
SFAd 16.073.5 8.970.2 9.170.2 8.770.1 8.770.2
MUFAd 13.272.8 21.270.4 19.370.5 19.570.2 19.670.2
Totaln-6PUFAd 4.972.1 7.170.2 7.370.2 7.370.1 7.270.1
Totaln-3PUFAd 0.770.2 1.170.0 2.570.1 2.670.0 2.670.0
C18:3n-3d 0.670.2 1.070.0 2.570.1 1.370.0 1.370.0
C20:5n-3d 0.0370.05 0.0170.00 0.0170.00 1.1270.02 0.1370.00
C22:6n-3d 0.0770.08 0.0270.00 0.0270.00 0.1770.00 1.0970.01
Dietaryfiber(g/MJ) 2.4770.92 2.7970.21 2.9670.16 2.9570.21 2.9670.22
VitaminE(mg/MJ)f 1.4570.55 3.6270.10 3.9970.11 5.1170.11 5.1770.09
Abbreviations:ALA,alpha-linolenicacid;DHA,docosahexaenoicacid;EPA,eicosapentaenoicacid;MUFA,monounsaturatedfattyacids;PUFA,polyunsaturatedfatty
acids;SFA,saturatedfattyacids.
aMean7s.d.
bCalculatedfrom3-daydietaryrecords.
cToconverttokcal/daymultiplyby239.
dValuesarepercentageofenergyintake.
eTotalfatcontainsB95%fattyacids,theotherB5%ismadeupofglycerolandotherlipids.
fa-Tocopherolequivalents.
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
317
Table 3 Fatty acid composition of the study oils used in the Inadditiontothestudyoils,thesubjectsreceivedspecial
experimentalperioda margarines as spread. Three different margarines were
produced (Unilever Health Institute). They contained
ALAoil EPAoil DHAoil
85% fat and 15% water. The total fat phase contained 85%
SSFA 7.71 7.08 7.12 of the respective study oil and 15% hard stock (palm oil,
C14:0 0.06 0.06 0.05 palmkernelfat).
C16:0 4.67 4.38 4.35
To compensate for possible variances in the individual
C16:1n-7 0.17 0.21 0.18
daily energy requirements, we developed special bread rolls
C18:0 1.93 1.58 1.58
SMUFA 59.89 60.68 60.97 with the same nutrientcomposition as the study diets.The
C18:1n-9 58.38 59.05 59.30 subjects got these rolls on request and could eat them ad
SPUFA 32.28 32.08 31.55
libitum without changing the composition of their diets.
C18:2n-6 22.90 22.35 22.36
Participants were supplied with food to meet 90% of their
C18:3n-3 8.91 3.56 3.59
C20:4n-6 0.0 0.23 0.1 daily energy requirements. The remaining 10% of energy
C20:5n-3 0.0 4.6 0.47 intake was taken up in the form of the so-called free-
C22:5n-3 0.0 0.03 0.35
choice foodstuffs that could be chosen from a given list.
C22:6n-3 0.0 0.58 4.22
VitaminE(mga-TE/kgoil)b 443.1 599.1 602.2 These foodstuffs (e.g. beverages, fruits, sweets) contained
only trace amounts of fat, tocopherol or protein. Subjects
Abbreviations: ALA, alpha-linolenic acid; DHA, docosahexaenoic acid; EPA, were instructed to keep their free-choice items constant
eicosapentaenoicacid;MUFA,monounsaturatedfattyacids;PUFA,polyunsa-
throughout the study. They recorded the food they had
turatedfattyacids;SFA,saturatedfattyacids.
a%oftotalfattyacids. chosenandtheconsumptionofextrarollsindiariesaswell
ba-TE,alpha-tocopherolequivalents. as any deviations from the diets. On the basis of these
records, we determined adherence to the diet, which was
were excluded from the diets. All diets contained identical foundtobeveryhigh.Theintakeofdrugsandanysignsof
basicfooditemsthatwerelowinfat,forexample,vegetables, illnesswerealsorecordedinthediaries.
whole-grainproducts,leanmeat,skimmedmilkandlow-fat
dairy products. This provided scope for the enrichment
ofthemealswiththestudyoilwhichwasincorporated,for
Laboratorymeasurements
example,insauces,desserts,curdandsaladdressings.
Blood sampling. All venous blood samples were taken
As basic study oil, we used a low-linolenic rapeseed oil
between06:30and08:15hafteranovernightfastofatleast
(3.7% ALA). For the three dietary groups, this oil was
12h under standardized conditions. Blood was drawn into
enriched either with ALA ethyl ester (ALA oil), EPA ethyl
tubeswithoutadditives,orcontainingethylenediaminetetra
ester(EPAoil)orDHAethylester(DHAoil)(Table3).Theoils
aceticacid(EDTA)orlithiumheparin(Sarstedt,Nu¨mbrecht,
were produced by Unilever Health Institute (Vlaardingen,
Germany).Plasmaandserumwasobtainedbycentrifugation
TheNetherlands).OwingtothenaturalALAcontentofour
at 2000g for 15min at 101C. After aliquotation in gas-tight
basicoil,allsubjectsingestedacertainamountofALAinthe
cryovials, plasma and serum were immediately frozen and
experimental period, and also in the wash-in period. This
storedat(cid:1)801C.Allbiochemicalanalyseswereconductedat
guaranteedaconstantbasiccontentofALAthroughoutthe
theendofthestudyinaseries.Allsamplesfromonesubject
entire study (1% of energy). To ensure that the observed
weremeasuredwithinoneanalyticrun.
effects could indeed be attributed to the enriched n-3 fatty
acids, the total fatty acid composition was kept constant.
Throughtheenrichmentoftheoils,thesubjectsadditionally Measurement of serum lipid parameters. Fasting serum con-
ingested 1.1% of energy as ALA, EPA or DHA. The mean centrations of total cholesterol and triglycerides were
daily intake of ALA in the ALA group was 6.071.1g (range measured using enzymatic colorimetric assays (Ro¨schlau
5.0–8.4g/day),themeandailyintakeofEPAintheEPAgroup et al., 1974; Siedel et al., 1993), high-density lipoprotein
was2.8g70.6g(range2.1–4.5g/day)andtheintakeofDHA (HDL) cholesterol and LDL cholesterol were directly
in the DHA group was 2.970.5g (range 2.2–3.8g/day). The measured by the methods of Sugiuchi et al. (1995, 1998).
ratioofn-6ton-3PUFAinallexperimentaldietswas2.8:1. AllmeasurementswereperformedinseriesonaHitachi917
In order to prevent dietary low cholesterol-PUFA-induced auto analyzer (Roche Diagnostics, Mannheim, Germany).
fatty acid oxidation in vivo, we added natural mixed The methods are validated by regular analyses of reference
tocopherols containing RRR-a-, RRR-b-, RRR-g- and RRR-d- serasuppliedbythenationalGermanINSTANDproficiency
tocopherols(Covi-ox T70)asantioxidants tothestudyoils. testing program and the international quality assurance
ThecontentofvitaminEneededwascalculatedonthebasis program of the US Centers for Disease Control and
oftheamountandtypeofPUFApresentintheoils.Foreach Prevention.Verylow-densitylipoprotein(VLDL)cholesterol
gram PUFA, expressed as equivalents of dienoic acid, an wascalculatedaccordingtothefollowingformula:
additional enrichment of 0.6mg a-tocopherol equivalents VLDLcholesterol(mmol/l)¼totalcholesterol(cid:1)
wascalculated(Muggli,1994). LDLcholesterol(cid:1)HDLcholesterol.
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
318
IsolationofLDL. LDLwasseparatedfromEDTAplasmaina (50m(cid:4)0.32mm i.d.(cid:4)0.2mm DF; Varian, Inc. Scientific
single run of 2h by density-gradient ultracentrifugation Instruments,Darmstadt,Germany),atemperature-program-
usingavariationofthemethodofChungetal.(1986).The mable vaporizer and a flame ionization detector (FID).
plasmawasadjustedwithsolidKBrtoadensityof1.26kg/l. Gas chromatograph conditions were: injector temperature:
Isotonic saline containing 0.1kg/l EDTA (purged with 60–2401C; oven temperature: 801C for 2min, increased in
nitrogen before use) was layered on top of the plasma in steps of 61C/min to 1401C and maintained for 4min, then
a Beckman Optiseal polyallomer ultracentrifuge tube increased in steps of 31C/min to 2251C and maintained for
(Beckman Coulter, Fullerton, CA, USA). The tubes were 6min; FID temperature: 2401C; flow rate: 1.8ml He/min;
centrifuged at 60000r.p.m. at 101C in a Beckman ultracen- solvent split 1:10. Results are expressed as mg fatty acid/mg
trifuge by using a Beckman near vertical rotor (NVT 65 LDLcholesterol.
Ti-Rotor).Afterultracentrifugation,theorangecoloredband
ofLDL,locatedinthetoplayerofthetube,wascollectedby
MeasurementofLDLa-tocopherolcontent. Theconcentration
usingasyringeandaneedleandfilteredthrougha0.22-mm
of a-tocopherol in LDL was determined by using reversed-
sterile filter (Renner, Dannstadt-Schauernheim, Germany)
phase high-performance liquid chromatography (HPLC).
intosterilevacuumcontainers(MallinckrodtRadio-Pharma,
Before analysis, 100ml of tocopheryl acetate (internal
Hennef, Germany). This preparation was stored in the dark
standard, 0.1mg/ml), 1ml of ethanol containing 0.1mg
at 41C. Susceptibility to oxidation was measured on the
BHT and 0.1mg EDTA and 20ml of a solution containing
following day. LDL for analyzing the LDL fatty acid
0.06M Na2WO4 and 1M MgCl2 in H2O were added to the
composition and LDL a-tocopherol content was stored at
LDL. Water (500ml) was added to the supernatant and the
(cid:1)801Cinopaquetubesuntilanalysis.
solution was extracted twice with 500ml n-hexane–dichlor-
methane–isopropanol (80:19:1, v/v/v). The pooled organic
Measurement of LDL susceptibility to oxidation. The ex vivo
phases were evaporated to dryness under nitrogen stream
oxidation of LDL was assessed using a modification of the
and the residue was then dissolved in 100ml ethanol
conjugateddiene(CD)assayasdevelopedbyEsterbaueretal.
and injected into the HPLC by means of a Rheodyne 7125
(1989). Briefly, LDL was desalted by gel-filtration on an
(Cotati, California, USA) 70ml loop injector. All extraction
Econo-Pac10DGcolumn(Bio-Rad,Munich,Germany)using
stepswereperformedoniceandindarkness.HPLCanalysis
phosphate-bufferedsaline(PBS)aselutingbufferandstored
was performed on a Kontron (Neufahrn, Germany) liquid
on ice for 60min until the oxidation was started. The
chromatograph (pump, model 422; column oven model
concentration of the desalted LDL solution was assessed by
480; UV diode array detector, model 440). Separation was
measurement of cholesterol content using a commercially
carried out on a 5mm C -resolve column (30cm(cid:4)3.9mm
18
available assay (CHOD-PAP, Roche Diagnostics, Mannheim, i.d.; Waters, Milford, MA, USA; column temperature, 301C)
Germany).ThedesaltedLDLwasdilutedto0.08gcholester-
and aflow rateof 1.2ml/min (solvent, acetonitrile–dichlor-
ol/linPBS.Theoxidationwasstartedat371Cbytheaddition
ethane–methanol (85:10:5, v/v/v)þ0.05% ammonium
of a freshly prepared CuSO solution (final concentration
4 acetate). Column eluates were monitored by ultraviolet-
1.6mmol/l)exactly1hafterdesalting.TheformationofCDs
visiblelightatwavelengthsof292nm.Valuesarequotedas
was monitored by continuous measurement of the change
a-tocopherol/mgLDLcholesterolforeachsample.
in absorbance at 234nm for 200min in a Kontron Uvikon
922 temperature-controlled spectrophotometer (Kontron,
Measurement of tocopherol equivalents in plasma. Concentra-
Neufahrn, Germany), resulting in a sigmoidal curve. A
tions of tocopherol equivalents in heparin plasma were
tangent to this curve was drawn at the point of inflexion.
determined by HPLC with UV-vis detection according to
Thelagtime(min)wasdefinedasthetimefromtheaddition
themethodofJakobandElmadfa(1995).Detectionofa-and
of CuSO until the intersection of this tangent with the
4
g-tocopherolswasperformedbyl¼295nm.Bothtocopherol
baseline. The rate of propagation (nmol CD/min/mg LDL
isomers were quantified by external standard calibration.
cholesterol) was calculated from the slope of the tangent,
Tocopherolequivalentswerecalculated,takingintoaccount
and the maximum amount of CD formation (nmol CD/mg
that the bioactivity of g-tocopherol in vivo is about 25% of
LDLcholesterol)wasdeterminedastheheightofmaximum
that of a-tocopherol (mg a-tocopherolþmg g-toco-
absorbanceabovebaseline.
pherol(cid:4)0.25¼mg tocopherol equivalents). The plasma
Measurement of total LDL fatty acid composition. The fatty tocopherol equivalent values are expressed as cholesterol
acid pattern of LDL was determined by gas-liquid chro- standardizedamounts(mgTE/mgtotalcholesterol).
matography using the method of Lepage and Roy (1986)
with the following minor modifications. C21:0 was used
as an internal standard. Methanol–benzene was replaced Statisticalanalysis
by methanol–toluol (4:1, v/v). Analysis was conducted All statistical analyses were performed using the SPSS
on a Trace gas chromatograph (Thermo-Finnigan, statistical software package (version 11, SPSS Inc., Chicago,
Bremen, Germany) equipped with a CP-Sil 88 column IL, USA). Serum concentrations of total cholesterol, LDL
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
319
Table4 Effectsofthestudydietsonfastingserumlipidandlipoproteinconcentrations
Variable Visita ALAdiet(n¼15) EPAdiet(n¼17) DHAdiet(n¼16)
Totalcholesterolb 1 5.0770.78 4.7770.72 4.9570.66
2 4.4270.67þþþ 4.3970.77þþþ 4.3770.74þþþ
3 4.3270.62 4.2570.87 4.4470.82
VLDLcholesterolb 1 0.7670.17 0.6470.17 0.7170.22
2 0.5870.14þþþ 0.5170.14þþþ 0.5870.24þþþ
3 0.5970.12 0.4970.15 0.6070.27
LDLcholesterolb 1 2.6270.62 2.4670.53 2.5670.57
2 2.1970.46þþþ 2.2270.50þþþ 2.1770.50þþþ
3 2.1470.43 2.1270.60 2.1770.56
HDLcholesterolb 1 1.6970.41 1.6770.30 1.6870.41
2 1.6570.42 1.6670.30 1.6270.36
3 1.5870.36 1.6570.30 1.6770.47
Triglyceridesc 1 1.16(2.17) 1.04(1.34) 1.23(1.40)
2 1.07(1.47)þþþ 0.77(0.97)þþþ 1.02(2.38)þþþ
3 1.26(1.46)*,w 0.78(0.98) 0.91(1.82)w
Totalcholesterol/HDLcholesterol 1 3.1370.80 2.9270.56 3.1270.92
2 2.7870.60þþþ 2.6870.45þþþ 2.8070.68þþþ
3 2.8370.62 2.6170.47 2.8170.80
Abbreviations:ALA,alpha-linolenicacid;DHA,docosahexaenoicacid;EPA,eicosapentaenoicacid;HDL,high-density lipoprotein; LDL,low-density lipoprotein;
VLDL,verylow-densitylipoprotein.
þþþPo0.001intragroupcomparisonsvisit2vsvisit1(pairedStudent’st-tests;fortriglycerides:Wilcoxon’ssigned-ranktests).
*Po0.05intragroupcomparisonsvisit3vsvisit2(Wilcoxon’ssigned-ranktest).
wPo0.05intergroupcomparisonsALAdietvsDHAdiet(Kruskall–WallisH-testfollowedbyMann–WhitneyU-test).
aVenousbloodsampleswereobtainedatthebeginningofthestudy(visit1),afterthewash-inperiod(visit2)andattheendofthe3-weekexperimentalperiod(visit3).
bmmol/l,mean7s.d.–toconvertcholesterolvaluestomg/dl,multiplyby38.66.
cmmol/l,median(range)–toconverttriglyceridestomg/dl,multiplyby88.57.
cholesterol, HDL cholesterol, VLDL cholesterol as well as Results
lag time, rate of propagation, maximum amount of CD
formation, LDL fatty acids and LDL a-tocopherol were all Effectofthedietsonserumlipidconcentrations
found to be approximately normally distributed, as con- Onthewash-indiet(visit1–2)therewasasignificantfallin
firmed by checking normal plots and histograms of the totalcholesterol(Po0.001),VLDLcholesterol(Po0.001)and
data and performing Kolmogorov-Smirnov tests and LDLcholesterol(Po0.001)aswellastriglycerides(Po0.001)
Shapiro–Wilk’s W-tests. In contrast, serum triglyceride (Table 4). The three study diets (visits 2–3) did not lead to
concentrations and plasma tocopherol concentrations ap- any significant changes in serum concentrations of total,
peared not to be compatible with a normal distribution. VLDL, LDL and HDL cholesterol (Table 4). Fasting concen-
Homogeneity of variances was verified by Levine’s test. trations of triglycerides did not change significantly after
Between-group comparisons were analyzed by using EPA or DHA diets. On the ALA diet serum triglycerides
analysis of variance (ANOVA) followed by post hoc slightlyincreased(Po0.05).
Tukey’s test if data were normally distributed and by
Kruskall–Wallis H-test and Mann–Whitney U-test if not.
EffectofthedietsonLDLsusceptibilitytooxidation
Within-group comparisons were performed using the
All three experimental diets (ALA, EPA, and DHA diet)
paired t-tests if data were normally distributed and
significantly decreased the lag time. The mean reductions in
analyzed by the Wilcoxon matched-pair signed-rank test if
theEPA(meandifference:(cid:1)9.64min((cid:1)16.4%),Po0.001)and
not. Unless otherwise indicated, results are expressed
DHA diet group (mean difference: (cid:1)9.51min ((cid:1)15.9%),
as mean7standard deviation (s.d.) or for skewed data
Po0.001) were significantly higher than in the ALA group
as median (range). All tests were two-tailed, and the
(mean difference: (cid:1)3.19min ((cid:1)5.1%), Po0.01) (Figure 1a).
Po0.05 level of significance was used to assess statistical
The EPA diet led to a decrease in propagation rate (mean
significance.
difference:(cid:1)0.37nmolCD/min/mgLDLcholesterol((cid:1)11.3%),
The coefficients of variation for all automated lipid
Po0.01) (Figure 1b), whereas the DHA diet increased the
measurements were below 5%, that of the measurement of
maximumamountofCDs(meandifference: þ7.85nmolCD/
LDL tocopherol was 2.8%, that of the measurement
mgLDLcholesterol(þ7.1%),Po0.001)(Figure1c).
of plasmatocopherol was 4.2%,those forthe measurement
of lag time, rate of propagation and maximum amount of
CD were 5.6, 6.0 and 6.4%, respectively, and those for EffectofthedietsonLDLfattyacidcomposition
the measurement of the LDL fatty acids were between 3.2 The experimental diets led to a significant enrichment of
and6.0%. LDL with the administered fatty acids ALA, EPA or DHA
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
320
(Table 5). However, the increase of ALA, EPA and DHA
differedbetweenthethreediets:ontheALAdiet,therewasa
1.9-fold increase in the ALA content (Po0.001). On the
EPA diet, the amount of EPA increased 9.1-fold (Po0.001)
and the DHA diet increased the DHA content by 3.0-fold
(Po0.001).
In the ALA diet, we also found an increase in the EPA
content(þ35.4%,Po0.01),butnosignificantchangeinthe
DHA content. In addition, there was a small decrease in
oleic acid (C18:1) (Po0.05) and arachidonic acid (C20:4)
(Po0.01). The EPA diet led to an additional enrichment
with ALA (þ19.8%, Po0.001) and with DHA (þ79.0%,
Po0.001), whereas the DHA diet further increased the
amount of ALA (þ20.0%, Po0.05) and EPA (þ284.2%,
Po0.001). On the latter, the relative increase in EPA (3.8-
fold)wassignificantlyhigherthantheincreaseinDHAitself
(3.0-fold). Neither on the EPA nor on the DHA diet, did we
observe significant changes in the oleic, linoleic or arachi-
donicacidcontents.
Effectofthedietsoncontentsofa-tocopherolinLDLand
contentsoftocopherolequivalentsinplasma
During the experimental period, the LDL a-tocopherol
content slightly decreased in the ALA group ((cid:1)0.82.mg/mg
LDL cholesterol ((cid:1)13.5%), Po0.05), and DHA group
(0.50mg/mg LDL cholesterol ((cid:1)7.3%), Po0.05). No signifi-
cant change was observed in the EPA diet group (P¼0.508)
(Table 6). The concentrations of tocopherol equivalents in
plasmaalsosignificantlydecreasedandtoasimilarextentin
all three experimental groups (ALA group: median change
(cid:1)5.0%(Po0.001),EPAgroup:(cid:1)5.7%(Po0.05),DHAgroup:
(cid:1)12.8% (Po0.05)). There were no significant differences in
the decreases in plasma tocopherol contents between the
threeexperimentalgroups(Table7).
Discussion
This strictly controlled dietary study was designed to
investigate the independent effects of ALA, EPA and DHA
on the ex vivo susceptibility of LDL to oxidation, the fatty
acid composition of LDL particles and the tocopherol
concentrationsinLDLandplasma.Then-3fattyacidswere
provided as ethyl esters, added to naturally low-linolenic
rapeseedoil.Thestudyoilswereincorporatedintodietswith
Figure1 Effectofthewash-indietandtheexperimentaldiets(ALA, normalfoodstuffs.
EPA, DHA diet) on lag time for CDs formation (a), rate of
The question whether dietary n-3 fatty acids render the
propagation (b) and maximum amount of CDs (c) (n¼48)
(mean7standard error of the mean). Venous blood samples were LDLparticlesmorepronetooxidativemodificationhasnot
obtained at the beginning of the study (visit 1), after the wash-in been conclusively answered until now. Oxidized LDL
period(visit2)andattheendoftheexperimentalperiod(visit3). particles are thought to be more atherogenic than native
þþþPo0.001 visit 2 vs visit 1 (intragroup comparisons, paired
particles (Steinberg et al., 1989; Witztum and Steinberg,
Student’st-test).*Po0.05visit3vsvisit2,**Po0.01visit3vsvisit2,
***Po0.001 visit 3 vs visit 2 (intragroup comparisons, paired 2001),and,consequently,theoftenreportedantiatherogenic
Student’s t-test). The significance levels for comparison between actionofn-3fattyacidscouldbepartiallyquenchedifthese
the diets areshownon the right,if Po0.05(ANOVA andposthoc fatty acids would enhance LDL oxidation. Our study shows
Tukey-test).
thatALA,EPAandDHAdifferintheireffectsonthedifferent
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
321
Table5 EffectsoftheexperimentaldietsonfattyacidcompositionofLDL
LDLfattyacids Visit ALAdiet(n¼15) EPAdiet(n¼17) DHAdiet(n¼16)
C18:1n-9a,b 2 191.96740.84 171.93739.98 193.41748.54
3 173.47745.25* 166.43738.77 175.57753.07
%change (cid:1)9.6713.6 (cid:1)1.7717.5 (cid:1)7.2721.9
C18:2n-6a,b 2 347.74755.51 314.25761.44 347.51799.02
3 338.94764.58 298.79734.17 328.37783.05
%change (cid:1)2.2713.5 (cid:1)1.7720.6 (cid:1)2.4721.5
C18:3n-3a,b 2 10.6274.62 9.5072.17 10.0372.11
3 19.4077.92***,w,z 11.1772.24***,w 11.8672.89*,z
%change þ88.9742.5 þ19.8720.0 þ20.0727.3
C20:4n-6a,b 2 67.43716.02 56.77712.80 61.35719.22
3 58.89716.26**,y 60.26712.02y 57.00718.94
%change (cid:1)12.3713.2 þ8.3719.8 (cid:1)4.8721.0
C20:5n-3a,b 2 9.6273.39 9.7073.69 9.3372.42
3 12.5574.12**,w,# 83.19729.59***,w,8 35.08711.73***,8
%change þ35.4733.0 þ808.97318.8 þ284.27135.3
C22:6n-3a,b 2 21.6275.55 19.9976.82 18.9676.07
3 20.0677.33z 33.3579.37***,z 50.36715.05***,z,z
%change (cid:1)4.0737.4 þ79.0761.2 þ199.87190.7
Abbreviations:ALA,alpha-linolenicacid;ANOVA,analysisofvariance;EPA,eicosapentaenoicacid;DHA,docosahexaenoicacid.
*Po0.05.
**Po0.01.
***Po0.001intra-groupcomparisonsvisit3vsvisit2(pairedStudent’sttests).
wPo0.001ALAdietvsEPAdiet.
zPo0.001ALAdietvsDHAdiet.
yPo0.01ALAdietvsEPAdiet.
#PoPo0.01ALAdietvsDHAdiet.
8Po0.001EPAdietvsDHAdiet.
zPo0.05EPAdietvsDHAdiet(ANOVAandposthocTukey-tests).
aMean7s.d.
bmg/mgLDL-cholesterol.
Table6 Effectsofthestudydietsona-tocopherolcontentofLDL
Variable Visita ALAdiet(n¼15) EPAdiet(n¼17) DHAdiet(n¼16)
LDLa-tocopherolb 1 4.6970.82 4.5470.68 4.8771.00
2 5.3771.10þþ 4.9570.76þþ 5.2871.17þþ
3 4.5570.87*,w 5.1271.06w 4.7870.78*
Abbreviations:ALA,alpha-linolenicacid;DHA,docosahexaenoicacid;ANOVA,analysisofvariance;EPA,eicosapentaenoicacid;LDL,low-densitylipoprotein.
þþPo0.01intragroupcomparisonsvisit2vsvisit1(pairedStudent’st-tests).
*Po0.05intragroupcomparisonsvisit3vsvisit2(pairedStudent’st-tests).
wPo0.05intergroupcomparisonsALAdietvsEPAdiet(ANOVAandposthocTukey-test).
aVenousbloodsampleswereobtainedatthebeginningofthestudy(visit1),afterthewash-inperiod(visit2)andattheendofthethree-weekexperimentalperiod
(visit3).
bmga-tocopherol/mgLDLcholesterol,mean7s.d.
LDLoxidationparameters.WiththeEPAdietweobserveda LDLrichinhighlyunsaturatedfattyacidsisnotnecessarily
reduction in lag time, indicating an enhanced oxidative oxidized more rapidly in biological systems than particles
susceptibility. This finding was confirmed by the investiga- containing fatty acids with fewer double bonds. The
tionsofMesaetal.(2004).However,incontrasttoMesaetal. mechanisms responsible for this phenomenon are not well
(2004), we simultaneously observed a reduced propagation understood, but several explanations have been postulated.
rate that indicates a decreased susceptibility to oxidation. Brudeetal.(1997)suggestthatEPAþDHAenrichmentofthe
This latter finding has also been reported in other human LDLparticleleadstoamoretightpackingofthelipidsthat
studies after ingestion of fish oils, which contained both, might makethedoublebondsmoreresistanttofreeradical
EPA and DHA (Suzukawa et al., 1995; Wander et al., 1996, attack, thus reducing the propagation rate. Wander et al.
1998;Sørensenetal.,1998;Foulonetal.,1999;Higdonetal., (1998)speculatethatprimarilytheconcentrationoflinoleic
2001; Pedersen et al., 2003). The slow propagation rate acid in LDL determines the propagation rate, both before
appears paradoxical and may support the hypothesis that and after EPA and DHA were introduced into the LDL
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
322
Table7 Effectsofthestudydietsonplasmatocopherolequivalentcontents
Variable Visita ALA-Diet(n¼15) EPA-Diet(n¼17) DHA-Diet(n¼16)
Tocopherollipid-adjustedb 1 4.35(3.05) 4.99(2.42) 5.09(3.20)
2 5.38(2.00)þþþ 5.25(3.33)þþþ 5.47(4.71)þþþ
3 5.11(2.21)*** 4.95(3.49)* 4.77(1.64)*
Abbreviations:ALA,alpha-linolenicacid;EPA,eicosapentaenoicacid;DHA,docosahexaenoicacid.
þþþPo0.05intragroupcomparisonsvisit2vsvisit1(Wilcoxon’ssigned-ranktests).
*Po0.05.
***Po0.001intragroupcomparisonsvisit3vsvisit2(Wilcoxon’s-signed-ranktests).
aVenousbloodsampleswereobtainedatthebeginningofthestudy(visit1),afterthewash-inperiod(visit2)andattheendofthethree-weekexperimentalperiod
(visit3).
bmg/mgtotalcholesterol,median(range).
particle.Theinfluenceoflinoleicacidover-rodethatofany thatthechangesinLDLoxidizabilitymighthavebeenmore
otherfattyacid.However,inourstudy,thelinoleiccontent pronounced if the MUFA content of the diets, and, subse-
inLDLremainedunchangedintheEPAdietaswellasinthe quently,theLDLhadbeenlowerbecauseMUFAarethought
ALA and DHA diets. As a third hypothesis, it has been tohaveantioxidantproperties(ReavenandWitztum,1996).
suggested that bicycloendoperoxides are formed in the This mayalsoexplain thatwe onlyfoundaminordecrease
presence of EPA and DHA and that they rapidly migrate to inlagtimeandnochangesinpropagationrateormaximum
the surface of the LDL particle because of their higher amount of CDs on the ALA diet. In accordance with the
polarity.Thus,becauseradicalsareremovedfromthesystem, results of Kratz et al. (2002), we speculate that a dietary
the rate of the propagation is slowed down (Higdon et al., intakeofALAofupto2.5%ofenergydoesnotexertadverse
2001).Inourstudy,theEPAdietincreasedtheLDLcontent affects with regard to LDL oxidation when provided in the
ofEPA,andalsothecontentsofALAandDHA.Therefore,we context of a diet rich in MUFA. This finding is of high
cannot conclude that the decrease in propagation rate is a practical relevance and supports current dietary guidelines
specificactiononlyofEPA.Furtherstudieswillbeneededto that recommend MUFA as main source of dietary fat,
investigate whether and how EPA influences LDL suscept- primarily provided with plant oils like rapeseed oil, which
ibilityinvitroandinvivoand,inaddition,todeterminethe isalsorichinALA(InternationalTaskForceforPreventionof
biological relevance of a reduced propagation rate and Coronary Heart Disease and International Atherosclerosis
concomitantdecreasedlagtimeinatherogenesis. Society,2003).
The DHA diet caused a marked decrease in the lag time. Anotherimportantfactorthatmustbetakenintoaccount
ThisfindingwasinagreementwiththeresultsofMesaetal. when discussing the LDL oxidizability is the tocopherol
(2004). However, they did not observe any changes in the content of diet and LDL particles. Evidence suggests that
amount of CDs, whereas we found a significant increase in lipid-soluble antioxidants, especially a-tocopherol, which is
this oxidation parameter. Both changes clearly indicate an presentinrelativelyhighconcentrationsintheLDLparticle,
enhanced LDL oxidizability. Especially the latter finding are directly protective against LDL oxidation (Esterbauer
mightbeofrelevancewithregardtothebiologicalactivityof etal.,1992).SupplementationofthedietwithvitaminEhas
oxidized LDL because the CD hydroperoxides are further been shown to increase the resistance of LDL to oxida-
decomposedtohighlyreactivesecondaryoxidationproducts tion (Reaven et al., 1993b; Jialal et al., 1995), and to
such as malondialdehyde or 4-hydroxynonenal (Esterbauer counteractthefishoil-inducedincreaseinLDLoxidizability
et al., 1992). This finding may indicate a potential athero- (Oostenbrug et al., 1994; Wander et al., 1996). However,
genicpropertyofDHA,althoughitmustbeemphasizedthat thesefindingswereobservedinstudieswithhighervitamin
it is an in vitro observation, and, based on these data, an E doses or combinations of vitamin E with other antiox-
in vivo progression of atherosclerotic processes that cannot idants(e.g.caroteneandvitaminC)(Brudeetal.,1997;het
bepostulated.Inaddition,itmustagainbenotedthatwith Hofetal.,1998;Upritchardetal.,2003)andmainlyaffected
theDHAdietwedidnotonlyfindanincreasedDHAcontent the lag time. In the present study, the LDL a-tocopherol
in LDL, but also substantial increases in ALA and EPA. contents and the concentrations of tocopherol equivalents
Therefore, it is difficult to determine whether DHA itself or inplasmaslightlydecreasedduringthe3-weekexperimental
theothern-3fattyacidswereresponsibleforthechangesin period.Thismayindicatethatincreasedoxidationoccurred
LDLoxidationparameters.Clinicalstudiesonthistopicare that would result in a larger consumption of antioxidants
highlywarranted. and,simultaneously,rendertheLDLparticleslessresistantto
The three n-3 PUFA ALA, EPA and DHAwere provided in oxidation. However, we found no significant correlations
thecontextofdietsrichinMUFA.ThisledtoLDLparticles between the tocopherol contents in LDL or plasma and
witharelativelyhighcontentofMUFA(about21.3%ofLDL theLDLlagtime(datanotshown).ThevitaminEcontentin
fattyacidsatvisit3),especiallyofoleicacid.Itcanbeargued the study fats was adapted to their contents in dienoic
EuropeanJournalofClinicalNutrition
N-3fattyacids,LDLoxidationandcomposition
SEgertetal
323
equivalents. There isno agreement on the exactamount of and DHA on the ex vivo LDL oxidation were somewhat
vitaminErequiredtocompensatefortheincreaseddemand conflicting,possiblyinpartduetotheircomplexmetabolic
caused by dietary PUFA in humans, especially for the pathways with a variety of possible conversion and retro-
long-chain and more unsaturated fatty acids EPA and DHA conversionreactions.Especiallywithregardtothenumerous
(Muggli,1994;ValkandHornstra,2000).Ourresultssuggest cardioprotective effects of n-3 fatty acids more studies
thatthestabilizationconceptweappliedwasappropriate. investigating them individually are needed to obtain more
TheenrichmentofthedietswithALA,EPAorDHAgreatly detailed knowledge of their specific metabolic actions
changed the n-3 fatty acid profiles in the LDL particles as an essential prerequisite for general and special dietary
confirming that the ingested fatty acids are effectively recommendations.
incorporated. On the ALA diet, we found a significant
increase in the LDL ALA content, and also an increase in
LDL content of EPA. Although our study design was not a
Acknowledgements
suitablemodelforinvestigatingconversionsofdietaryfatty
acids, we speculate from our results that ALA might have
WeareindebtedtoAGolla,EDeppe,JGrimme,JHameister,
been converted to EPA through a series of elongation and
S Irmer, S Kocksch, P Mahlsta¨dt for their excellent work on
desaturation reactions. This conversion has been shown
thedietarystudy,andWHanekamp,SWentkerforexcellent
in several metabolic studies (Pawlosky et al., 2001; Burdge
technicalassistance;toGBerger,SVenhermandMEisenberg
andWootton,2002;Burdgeetal.,2002,2003;Goyensetal.,
for performing the venipunctures; to M Kratz for helpful
2005;Husseinetal.,2005).Inagreementwiththeresultsof
comments in planning the study; to Unilever Research
most human intervention trials that analyzed fatty acids in
Vlaardingenforproductionofthestudyoilsandmargarines;
lipoproteins, plasma or blood cells, in our study a dietary
to M Delker and W Bernhard at the Bildungszentrum der
ALAintakeatdosesofabout6g/dayor2.5%ofenergyfailed
Bundesfinanzverwaltung for their generous cooperation.
toincreaseDHAconcentrations(Kratzetal.,2002;Finnegan
Last but notleastwe thank thevolunteers for theirinterest
etal.,2003;Husseinetal.,2005;Harperetal.,2006).Basedon
andparticipationinourstudy.
the studies of Burdge et al. (2002, 2003) and Hussein et al.
(2005) one could assume that ALA cannot be substantially
converted to DHA in humans, although a few studies
References
found a very limited conversion efficiency of dietary ALA
to DHA (Emken et al., 1994; Pawlosky et al., 2001; Goyens
BonanomeA,BiasiaF,DeLucaM,MunarettoG,BiffantiS,PradellaM
etal.,2005).
etal.(1996).N-3fattyacidsdonotenhanceLDLsusceptibilityto
With the EPA diet, we found a significant increase in the oxidationinhypertriacylglycerolemichemodialyzedsubjects.Am
EPA content of LDL particles, but also an increase in JClinNutr63,261–266.
DHA content. This may suggest a conversion of EPA to Brossard N, Croset M, Pachiaudi C, Riou JP, Tayot JL, Lagarde M
(1996). Retroconversion and metabolism of (13C)22:6n-3 in
DHA, as documented previously (Burdge and Calder,
humans and rats after intake of a single dose of (13C)22:6n-3-
2005). Another important finding is that the EPA enrich- triacylglycerols.AmJClinNutr64,577–586.
ment of the LDL particles on the EPA diet was more BrudeIR,DrevonCA,HjermannI,SeljeflotI,Lund-KatzS,SaaremK
pronounced than the enrichment with equal amounts of etal.(1997).PeroxidationofLDLfromcombined-hyperlipidemic
malesmokerssuppliedwithomega-3fattyacidsandantioxidants.
DHA on the DHA diet (ninefold vs. threefold; Table 5).
ArteriosclerThrombVascBiol17,2576–2588.
SowespeculatethatEPAisincorporatedintheLDLparticles
Burdge G, Calder PC (2005). Alpha-linolenic acid metabolism in
withparticularpreference. adult humans: the effects of gender and age on conversion to
The DHA diet did not only significantly increased the longer-chain polyunsaturated fatty acids. Eur J Lipid Sci Technol
107,426–439.
DHA content of LDL particles, but also their EPA content,
BurdgeGC,FinneganYE,MinihaneAM,WilliamsCM,WoottonSA
suggestingaretroconversionofthelonger22:6fattyacidto (2003).Effectofaltereddietaryn-3fattyacidintakeuponplasma
its 20:5 derivative. Such a retroconversion has been docu- lipid fatty acid composition, conversion of (13C)alpha-linolenic
mented in several supplementation studies with purified acid to longer-chain fatty acids and partitioning towards beta-
oxidationinoldermen.BrJNutr90,311–321.
DHA(ConquerandHolub,1996,1997;Vidgrenetal.,1997;
Burdge GC, Jones AE, Wootton SA (2002). Eicosapentaenoic
Hansen et al., 1998) and in one study of DHA metabolism
anddocosapentaenoic acidsare the principal products of alpha-
using[13C]DHAinhumans(Brossardetal.,1996). linolenicacidmetabolisminyoungmen.BrJNutr88,355–363.
In summary, an increased dietary intake of ALA, EPA or BurdgeGC,WoottonSA(2002).Conversionofalpha-linolenicacid
to eicosapentaenoic, docosapentaenoic and docosahexaenoic
DHA led to a significant enrichment of the respective fatty
acidsinyoungwomen.BrJNutr88,411–420.
acid in the LDL particles. In addition, dietary ALA also
ChungBH,SegrestJP,RayMJ,BrunzellJD,HokansonJE,KraussRM
caused an EPA enrichment. Dietary EPA seemed to be etal.(1986).Singleverticalspindensitygradientultracentrifuga-
preferentially incorporated in the LDL particles, and, tion.MethodsEnzymol128,181–209.
ConquerJA,HolubBJ(1996).Supplementationwithanalgaesource
simultaneouslyincreasedtheirDHAcontent.Inthecontext
of docosahexaenoic acid increases (n-3) fatty acid status and
of aMUFA-richdiet, we conclude thatALA enrichment did
altersselectedriskfactorsforheartdiseaseinvegetariansubjects.
not enhance LDL oxidizability, whereas the effects of EPA JNutr126,3032–3039.
EuropeanJournalofClinicalNutrition