Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
1
SpecialArticle1
Evaluating and adapting the Mediterranean diet for nonMediterranean2
populationsacriticalappraisal3
4
Authors5
R.Hoffman(correspondingauthor)6
SchoolofLifeandMedicalSciences,UniversityofHertfordshire,Hatfield,AL107
9AB,UK8
Tel.01707284526;Fax017072850469
E‐Mail:[email protected]
11
M.Gerber12
CancerResearchCenter,INSERM‐CRLC,34298,Montpellier,cedex5,France13
14
Keywords15
Mediterraneandiet;Mediterraneandietscore;confoundingfactors16
17
Abstract18
This review discusses the limitations of current techniques for evaluating the19
Mediterranean diet in non‐Mediterranean populations, and how differences20
between theeatingand lifestylehabitsof these twopopulationsmay influence21
2
the implementationofaMediterraneandiet innon‐Mediterraneanpopulations.22
Food groups may vary significantly between Mediterranean and non‐23
Mediterraneanpopulationsdue to theprecise foodswithin the foodgroupand24
due to aspects of foodproduction andpreparation.Examples includeMUFA in25
relationtoitssource(meatversusoliveoil),vegetablesinrelationtotheamount26
consumedandhowtheyareprepared,alcohol in termsof itspatternof intake27
and source (wine versus other sources), and the nature of meat and dairy28
produce. Lifestyle factors such asmeal patterns and exposure to sunlightmay29
also act as confounding factors when assessing the overall benefits of a30
Mediterraneandiet.ImprovingMediterraneandietscoresandmeasuringplasma31
nutrient levels may help mitigate the effects of these confounders. These32
considerations may have important health implications when implementing a33
Mediterraneandietinnon‐Mediterraneanpopulations.34
35
3
35
Introduction......................................................................................................................................336
DietaryassessmentsofadherencetoaMeddiet .............................................................637
Aprioridietarypatterns ........................................................................................................638
CorrelatingandconfoundingfactorsassociatedwithaMeddiet..................... 1039
Comparisonbetween foodsand theirpreparationmethods inMediterranean40
andnon‐Mediterraneancountries....................................................................................... 1141
Oliveoil ....................................................................................................................................... 1242
Vegetables.................................................................................................................................. 1243
FruitsandNuts ........................................................................................................................ 1644
Cereals......................................................................................................................................... 1945
Legumes...................................................................................................................................... 1946
Fish ............................................................................................................................................... 2047
Dairyproduce .......................................................................................................................... 2248
Meat.............................................................................................................................................. 2449
Otherfoods................................................................................................................................ 2550
Alcohol ........................................................................................................................................ 2651
Discussion....................................................................................................................................... 2852
Conclusions.................................................................................................................................... 3153
Acknowledgements .................................................................................................................... 3154
55
Introduction56
DespitethemultipleformsoftheMediterraneandiet(Meddiet)thathavearisen57
asaconsequenceofthediversefoodhabitsacrosstheregion,nutritionistshave58
neverthelessestablishedamodelMeddiet.ThismodelMeddietischaracterised59
byalargequantityanddiversityofplant‐derivedfoods(wholegraincereals,raw60
andcookedvegetables,freshanddryfruits,legumesandnuts),fish,amoderate61
intakeofmeatanddairyproduce(bothpreferablyfromgoatsandsheep),olive62
oil as added fat, and amoderate intake of wine duringmeals. 1,2 This dietary63
4
pattern typically represented the food habits of Southern Italy and Greece64
(especially Crete) around the 1970s. These countries were part of the “Seven65
CountriesStudy”,whichwasthefirststudytodemonstratethehealthbenefitsof66
a Med diet with regard to all‐cause mortality, and especially cardiovascular67
mortality, when comparing a traditional Med diet with diets from the US and68
NorthEuropeancountries.3Later,thehealtheffectsoftheMeddietwerestudied69
within a single population by classifying the subjects into groups according to70
theiradherencetoaMeddiet, 4andbyevaluating theirdiseaseriskrelativeto71
how far their food habits were from the traditional Med diet. 5 Both of these72
studies were conducted within Mediterranean populations and used a priori73
dietarypatternsthatwerebasedonthemaincharacteristicsoftheMeddiet.74
75
Subsequently, itwas proposed to applyMeddiet scores to non‐Mediterranean76
populations.However,thereareanumberofdifferencesbetweentheeatingand77
lifestylehabitsofMediterraneanandnon‐Mediterraneanpopulationsthatmake78
usingtheoriginalMeddietscoreseitherimpossible(forexample,whentheMed79
dietscoreincludesoliveoil,sinceonlyafewpopulationsintheworlduseolive80
oil), or misleading (for example if the foods of the score have different81
compositions in Mediterranean and non‐Mediterranean countries). This latter82
pointcouldpotentiallyinfluencetheevaluationoftheeffectivenessofaMeddiet83
fornon‐Mediterraneanpopulations.Analysingtheresultsofvariousstudiesthat84
haveappliedaMeddietscore(eithertheoriginalone,oraderivedormodified85
score),itispossibletopinpointthosefeaturesoftheMeddietthatappearmost86
stringent.This is importantbecauseof the interest inusingaMeddiet innon‐87
5
Mediterraneanpopulations tomanagecoronaryheartdisease (CHD)andother88
chronicdiseases.6,789
90
This reviewdiscusses someof thedifferencesbetween the eating and lifestyle91
habits of Mediterranean and non‐Mediterranean populations that could92
potentially impacton thehealthbenefitsofaMeddiet.Weaddress threemain93
issues:94
1. Are the a priori Med diet scores that have been constructed from95
literature data or dietary assessment obtained in Mediterranean96
populations appropriate for evaluating adherence to aMed diet and its97
healtheffectsinnon‐Mediterraneancountries?98
2. Arethere factorsassociatedwithaMeddiet inMediterraneancountries99
butnotpresentinnon‐Mediterraneancountriesthatmayaffectthehealth100
benefits obtained from aMed diet (eg differences in lifestyle andmeal101
structure)?102
3. Coulddifferencesinthetypesandvarietiesoffoodsconsumedandtheir103
methodsofpreparationalsoplayarole?104
105
Theseissueswillbeillustratedbycomparingthewesterndiet,andespeciallythe106
UKdiet,withthedietinMediterraneancountries.107
108
6
DietaryassessmentsofadherencetoaMeddiet109
Aprioridietarypatterns110
By the 1990s, nutritional epidemiology had foreseen the limits of using an111
assessment system based on nutrient intake. The ability to understand the112
relationshipbetweenfoodandhealthwasrestrictedbecauseofincompletefood113
compositiontablesandbecauseofthedifficultyinascertainingthespecificityof114
a given nutrient, or even a given food, to an observed health effect. It became115
evident that aholistic approach shouldbeused innutritional epidemiology, 8,9116
and this especially applied to theMed diet, a dietary patternwhose beneficial117
effects were becoming widely acknowledged. 10 Several Med diet scores were118
thereforedesigned.5,11,12OnlytheMeddietscoresofTrichopoulouetal,referred119
tohereasMDS,willbediscussedherebecause it is theonemostwidelyused,120
andalsobecauseithasbeenmodifiedtofitvariouspopulations.121
122
IntheoriginalMDS,theauthorstookadvantageofthefoodconsumptionpattern123
of an elderly Greek populationwho hadmaintained a traditionalMed diet up124
untilthe1990s.5Anutritionalsurveyofthispopulationprovidedaquantitative125
assessment of the characteristic foods of the Med diet. A point scale was126
established to assess the degree of adherence to the traditionalMed diet. The127
authors arbitrarily decided that the median sex‐specific value of the128
consumptionofnineselectedcharacteristiccomponentsof theMeddietwould129
bethecut‐offdeterminingadherence.Avalueof0or1wasassignedforeachof130
thefivepresumedbeneficialfoodgroupsievegetables,legumes,fruitsandnuts,131
cerealsandfish.Avalueof0wasassignedifconsumptionwasbelowthemedian,132
7
and a value of 1 was assigned if consumption was above the median. It was133
decided to replace olive oil with the ratio of monounsaturated fatty acids134
(MUFAs)tosaturatedfattyacids(SFAs)becauseAncelKeyshadestablishedthat135
this ratio was the main factor contributing to the low cardiovascular disease136
(CVD)mortality.3Thusavalueof1wasassignedtoindividualswithaMUFA:SFA137
ratioatorabovethemedianobserved inthepopulation,andavalueof0 fora138
ratio less than the median. For dairy products, and meat and meat products139
(whoseconsumptionistypically lowormoderateintheMeddiet),avalueof0140
was assigned for consumption at or above the median, and a value of 1 was141
assignedforconsumptionbelowthemedian.Withregardtothemoderateintake142
ofethanolintheMeddiet,avalueof1wasassignedtomenconsumingbetween143
10gand<50gperdayandtowomenconsumingbetween5gand<25gper144
day. Thus, the final MDS ranged from 0 (minimal adherence) to 9 (maximal145
adherence).146
147
ThispatternwasperfectlyadaptedtotheGreekpopulationandconformedwith148
previous results on the benefits of aMed diet onmortality. 5 Itwas therefore149
proposed to apply it to other populations, e.g. populations in the EPIC study.150
However,afewmodificationswerenecessaryinordertoadapttheMDStoother151
Europeanpopulations.ThescorefortheEPICstudyhadtwomajoradaptations152
13:153
(1) In theoriginal score, themajor sourceofMUFAs isoliveoil and the154
high MUFA:SFA ratio reflects a high consumption of olive oil and a low155
consumption of animal products. Because relatively few people in Northern156
8
Europeuseoliveoil,itwasnecessarytomodifythescoreusinganothermarker157
for vegetable oil consumption.Hence theMUFA:SFA ratio becameunsaturated158
fat:SFA. This had two consequences. Firstly, it did not take into consideration159
substancesinthenon‐saponifiablefractionofoliveoil(especiallythetriterpene160
squaleneandthephenoliccompoundshydroxytyrosolandoleocanthal),manyof161
which have antioxidant and anti‐inflammatory actions and are potentially162
beneficialagainstchronicdegenerativediseases,andhencemortality.Secondly,163
MUFAsare,toalargeextent,markersofoliveoilintheoriginalMDSsinceitwas164
applied to a populationmainly consuming olive oil for cooking and as a salad165
dressing.ButwhenappliedtoNorthernEuropeanpopulations,MUFAsreflect,to166
alargeextent,theconsumptionofanimalfat,andithasbeendemonstratedthat167
these MUFAs are associated with CVD 14 and breast cancer, 15 as further168
discussed by Gerber and Richardson. 16 It is interesting to note that in an “a169
posteriori”adaptationoftheMDS,oliveoilreplacedtheratioMUFA:SFA.17170
(2) A significant difference between Greek and North European171
populationsisintheirintakeofplant‐derivedfoods.RetainingtheGreeklevelof172
consumption would have resulted in the majority of North European people173
scoring0forvegetablesandlegumes.Hence,itwasdecidedtouseasacut‐offthe174
median of the consumption of the nine components of the MDS in each175
population. Under these conditions, the value of the cut‐off differedmarkedly176
among the EPIC populations. For example, the median Greek consumption of177
vegetables(excludingpotatoes)intheEPICstudywas500g/dformenand550178
g/dforwomen,5whereasthemedianvaluesforallpopulationsoftheEPICstudy179
were157g/dayformenand184g/dforwomen.13Asmightbeanticipated,such180
amodifiedMDSwasnot associatedwitha reducedmortality risk in theNorth181
9
Europeanpopulationsanalysed.13TheaposterioriadaptationoftheMDSofSofi182
etalproposedabsolutevalues for foodconsumption tobe reached inorder to183
describe adherence to a Med diet. 17 These values were derived from a184
segmentedregressionanalysisofeachfoodconsumptionoftheMDSandoverall185
mortalityoftheGreekEPICcohort.The“change‐point”onthesegmentprovided186
forthefoodcut‐offtobeusedintheMDS.Thusforvegetablesthe“change‐point“187
is >500 g, very close to what could be deduced from the literature, and the188
amountthatisgenerallyproposedinpublichealthrecommendations.189
190
SeveralothermodifiedMeddietscoreshavesubsequentlybeenproposed.Inthe191
US, Fung and colleagues excluded potato products from the vegetable group,192
separated fruit and nuts into two groups, eliminated the dairy group because193
low‐fat milk is predominantly used in the US, included whole‐grain products194
only,includedonlyredandprocessedmeatsinthemeatgroup,andallocated1195
pointforalcoholintakebetween5and15g/d.18ItwascalledthealternateMDS196
(aMDS)andwasnotassociatedwithoverallpost‐menopausalbreastcancerrisk,197
but only in sub‐classes of this disease. In 2010, another modification was198
developed and applied to the EPIC cohorts. 19 Each component (apart from199
alcohol)wascalculatedasafunctionofenergydensity(g/1000Kcal/d)andwas200
thendividedintotertilesofintake.Ascoreof0,1or2wasassignedtothefirst,201
second,andthirdtertilesof intake for the fivecomponentspresumedto fit the202
Med diet, namely fruit (including nuts and seeds), vegetables (excluding203
potatoes), legumes, fish(freshorfrozen,excludingfishproductsandpreserved204
fish), andcereals.Thescoringwas inverted for the twocomponentspresumed205
10
not to fit the Med diet, namely total meat and dairy produce. This scoring206
recognizedthe importanceofoliveoil:0wasassignedtonon‐consumers,1 for207
subjects consuming below the median (calculated only within olive oil208
consumers), and2 for subjects consumingequalorabove thismedian.Alcohol209
wasassignedeither2formoderateconsumers(range:5–25g/dforwomenand210
10–50g/d formen)or0 forsubjectsoutside(aboveorbelow) thesex‐specific211
range. This modification was called the relative Med diet (rMED). The rMED212
scoreswerethengroupedintolow(0–6),medium(7–10),andhigh(11–18).This213
scorecompensatedforthedifferenceinfoodconsumptionamongthecohortsby214
usingenergydensityandtookintoconsiderationthespecificityofoliveoil.The215
resultsarenotshownbycohorts,buttheoverallresultsshowedariskreduction216
forgastricadenocarcinomaof33%withincreasingadherencetotherMED.217
218
CorrelatingandconfoundingfactorsassociatedwithaMeddiet219
Ithasbeenshownthattheorderofcourses inamealandthepatternofmeals220
through the day are strong characteristics of aMed diet. 20 Lunch is themain221
meal, providing not only the required energy but also a sufficient quantity of222
plant‐derivednutrients,iefibre,micronutrientsandphytochemicals.Inaddition,223
besides the expected typical food intake, sharing of meals with family or224
colleagues and an absence of snacking were found to be major features in225
Sardinia,wherethereishighadherencetoatraditionalMeddiet,comparedwith226
Malta, where many traditional features have been lost. 20 Although these227
differentfactorsmightplayanadditionalroletofoodcomponentsinpreserving228
11
health, they are not currently assessed in dietary questionnaires and hence229
cannotbecontrolledforbystatisticaltechniques.230
231
Severalotherfactorsmaybereducedorabsentinthelifestylesofpeoplelivingin232
non‐Mediterranean countries. A healthy energy balance derived from physical233
activity was an intrinsic part of the Med diet in Mediterranean populations234
following a traditional lifestyle. Also, taking a siesta ‐ still current in some235
Mediterraneancountries ‐hasbeenassociatedwitha lowerriskofCVD, 21but236
this is rarely considered in epidemiological studies. Another potentially237
confounding factor in relation to several chronic degenerative disorders is the238
possibleprotectiveeffectsconferredbyvitaminD.22,23Becauseofmoreintense239
sunlight, the UV‐induced synthesis of vitamin D will supplement nutritional240
intaketoagreaterextentformanypeopleinMediterraneancountries,compared241
tomorenortherlycountries.242
243
Comparison between foods and their preparation methods in244
MediterraneanandnonMediterraneancountries245
246
The precise composition of a food and how it is produced and prepared can247
differ significantly between Mediterranean and non‐Mediterranean countries.248
ThesedifferencesmayimpactontheoverallhealthbenefitsofaMeddietinnon‐249
Mediterraneancountries.250
251
12
Oliveoil252
ThemajordifferencebetweennorthernandMediterraneancountriesinrelation253
tooliveoilissimplywhetherornotitisconsumed,andthishasbeendiscussed254
earlier. However, the quality of olive oil and its culinary use may also be255
importantwith regard to its health benefits. AlthoughMUFA content does not256
varysignificantlybetweendifferentqualitiesofoliveoil,oliveoilalsocontainsa257
"non‐saponifiable" fraction comprising various triterpenes (mostly squalene),258
phytosterols(mostlyβ‐sitosterol),tocopherols(mostlyvitaminE)andphenolics,259
and these may vary between various types of olive oil and be influenced by260
culinarypractices.Levelsofsqualene,aputativeprotectivefactoragainstbreast261
cancer,24phytosterolsandtocopherolsarereducedwithincreasingrefinement262
ofvirginoliveoil.Inaddition,phenolicsareonlypresentinsignificantquantities263
invirginoliveoilandsotheirhealthbenefitswillbelostwhennon‐virginolive264
oils are consumed. Potentially important phenolics in virgin olive oil include265
lignans, which are associated with reduced breast cancer risk,25,26266
hydroxytyrosol, which has cardioprotective and anticancer activity in267
experimental systems, 27 and the anti‐inflammatory substance oleocanthal.268
Fryingwithvirginoliveoil reduces thephenolic content, 28 and thishighlights269
thepotentiallyimportantrolethatconsumptionofrawvirginoliveoilmayplay270
intheMeddiet.271
272
Vegetables273
Populations inMediterraneanandnon‐Mediterraneancountriescanhavequite274
differentpreferencesfortypesofvegetablesandtheirpreparationmethods,and275
13
this may influence health outcomes in relation to vegetable consumption. 29276
Dutchuniversitystudentswerereportedtoprefer"Brusselssprouts,greenpeas277
andcarrotsoftenwithapplesauce"whereasGreekstudentschose"mostlyfresh278
saladsoftomatoes,cucumber,cabbage,rocket,radishes,spinachandlettucewith279
olive oil, vinegar or balsamic vinegar and herbs, or green vegetables in baked280
pies (spinachpie, leakpie, etc.)". 30 Saladswere found tobe consumedseveral281
timesaweekinananalysisofthetraditionalCretanMeddiet.31InEPICcohorts282
fromnortherncountries,consumptionofrawvegetablesasaproportionoftotal283
vegetable consumption was reported to be lower compared to their southern284
counterparts. 32Thiswasparticularly striking forUKmenwhoconsumedonly285
half asmuch rawvegetables as cookedvegetables,whereas theproportionsof286
rawandcookedvegetableswerefairlysimilarinMediterraneancountries.Data287
fromtheUKLivingCostsandFoodSurveyindicatethatthemainpurchasedfresh288
vegetables in the UK in 2010 included cabbages, brussel sprouts, cauliflower,289
saladleaves,carrots,alliumsandtomatoes.33290
291
Thesewide variations in preferred types of vegetablesmay have a significant292
impactonphytochemical intake andhenceonany correlateddisease risk.The293
lowconsumptionofdarkgreenleafyvegetables(egbroccoli,spinach,kale)inthe294
UK32 isnoteworthysincetheserepresentamajordietarysourceofvitaminsC295
andK,folate,β‐carotene,lutein+zeaxanthinandflavones.34Garlicconsumption,296
commoninMediterraneancountries,islowintheUK,andtheWCRF/AICR2007297
reportconsideredit"probable"thatgarlicconsumptioncontributestoprotective298
effects against stomach and colorectal cancers. 35 Differences in the299
14
phytochemicalcontentbetweendifferentvarietiesofthesamevegetablecanbe300
substantial. For example, flavonol content of lettuce varieties ranged from 0.5301
µg/g freshweight for iceberg lettuce ‐avarietycommonlypurchased inNorth302
Europeancountries‐to207µg/gfreshweightfortheItalianvarietylollorosso.303
36DietaryflavonolintakeislinkedtoadecreasedriskofCVD,includingstroke.37304
Of course, low consumption of one vegetable may be compensated for by305
consumptionofanothervegetablecontainingthesamebeneficialnutrients,and306
moreovermany phytochemicals present in vegetables are also found in fruits.307
Thus, amixed and diverse diet can help ensure an optimum intake of a wide308
rangeofhealthyphytochemicals.Thepotential importanceofavarieddietwas309
highlightedinanEPICstudywhichshowedthatdiversityoffruitandvegetables310
consumptionwasassociatedwithadecreasedriskoflungcancer,aneffectover311
and above the inverse association with quantity. 38 It is noteworthy that a312
traditional Med diet includes a particularly wide diversity of fruits and313
vegetables.314
315
Preparation method can influence both nutrient levels and nutrient316
bioavailability. Consumption of raw vegetables preserves heat labile nutrients317
suchasvitaminsAandCandfolatesthatotherwisecanbelostwhenvegetables318
are cooked. Also, using an oil‐based dressing ‐ olive oil is traditional in319
Mediterranean countries ‐ was shown to increase the bioavailability of320
carotenoidsfromsaladingredients.39Breakingdownthefoodmatrixbycooking321
orpureeingcanalso increase thebioavailabilityof carotenoids.40Thecommon322
practice inNorth European countries of boiling a single type of vegetable and323
15
discarding the cooking water can result in significant nutrient loss due to324
leaching ofwater soluble nutrients such as folates and glucosinolates into the325
cooking water. 41,42 This practice is less common in Mediterranean countries,326
where soups and stews arepreferred, and since these cookingmethods retain327
the cooking medium, there is no loss of water soluble nutrients. Frying328
vegetables can lead to significant losses of fat soluble nutrients such as329
carotenoids, probably into the cooking fat, 43 but this practice is not very330
commonintheMeddiet.Ontheotherhandwhentheentirecontentsofthepan331
areconsumed,forexampleinMediterraneanstews,theoverallnutritionalvalue332
ofthedishwillnotbecompromised.333
334
The emphasis in a traditionalMeddiet is for seasonal, field‐grown vegetables,335
whereas for a North European market "Mediterranean" vegetables are336
frequently grown under glass. This latter cultivation practice reduces UV‐B337
exposuredue to the limited light absorbingproperties of glass. It is also often338
accompaniedbyincreasedfertiliseruse.Bothofthesefactorscanhaveadverse339
effects on phytochemical production, although this appears to dependboth on340
thetypeofvegetableandthespecificphytochemical.44341
342
In conclusion, the nutritional benefits that different populations receive from343
consuming vegetables may vary widely, and is not accounted for by simply344
determiningabsolutelevelsofconsumption.345
346
16
FruitsandNuts347
TotalfruitconsumptionintheUKislowcomparedtoMediterraneancountries,348
and across Europe there is a North South gradient for total consumption. 32349
Apples,bananasandcitrusfruitstogetheraccountedforabouttwo‐thirdsofall350
fresh fruitpurchased in2010byUKhouseholds,andprocessed fruitsand fruit351
products(excludingfruitjuices)accountedforathirdoftotalfruitpurchases.33352
Consumption of summer fruits popular in Mediterranean countries ‐ such as353
pomegranates,figs,grapesand"orangefruits"(egapricots,peaches,nectarines,354
cantaloupemelons)‐isrelativelylowintheUK.355
356
When eaten raw,many fruits are a good source of vitamin C. By contrast, the357
relative amounts of various phytochemicals can vary widely: citrus fruits are358
good sources of flavanones and flavones, berries are rich in anthocyanins and359
flavan‐3‐ols, and Mediterranean "orange" fruits are important sources of α‐360
carotene and β‐carotene. Due to the difficulty for epidemiological studies to361
identify the effects of individual nutrients within a diet, it is unclear if the362
particularnutrientcontentofaspecific fruitaffectshealthoutcomes.However,363
Chong et al concluded that therewas some limited evidence that fruits rich in364
flavonols,anthocyaninsandprocyanindins,suchaspomegranate,purplegrapes365
andberries,aremoreeffectiveatreducingCVDrisk.45366
367
Fruit isatypicalwaytoendamealinMediterraneancountries.Attheendofa368
meal there is a rise inpro‐oxidant andpro‐inflammatoryprocesses and this is369
linked to increased cardiovascular damage. Postprandial hyperlipidaemia and370
17
hyperglycaemia are also risk factors for a number of metabolic disorders371
including type 2 diabetes, CVD and metabolic syndrome. Some studies have372
shown that consumption of phenolic‐rich fruits during the postprandial phase373
increases the antioxidant capacity of the blood46Hence consuming fruit at the374
endofamealisaprudentstrategyforhealthyeating.375
376
In conclusion, consumption of awide range of fruits is advisable, especially if377
phytochemical intake from vegetables is limited. Eating fruit, rather than a378
pastry, at theendof amealnotonly reduces calorie intakebut fruitphenolics379
mayalsohelpcounteractoxidativestressandotherpathologicaleventsduring380
thepostprandialphase.381
382
AnanalysisofnutconsumptionintheEPICstudyfoundthatcohortsfromcentral383
European countries (North of France, Germany, Netherlands, UK) and384
Mediterranean countries (South of France, Greece, Italy, Spain) had similar385
overall levels of consumption. 47 However, a higher proportion of peanuts386
relativetotreenuts(mainlywalnuts,almondsandhazelnuts)wereconsumedin387
the central European countries. For example in the UK, peanuts and tree nuts388
constituted 40.4% and 36.4% respectively of total nuts and seeds consumed,389
whereasinSpainthesefigureswere26.8%and54.9%respectively.390
391
AlltypesofnutshavebeenshowntoreducetheriskofCHD,althoughonlyvery392
limiteddataisavailableforpeanuts.48Nutshavehypocholesterolaemiceffects,393
andanumberof intervention studieshavedemonstrated thatnuts lowerboth394
18
LDL‐cholesterol and the ratio of LDL‐cholesterol to HDL‐cholesterol. 48395
Participants of the PREDIMED study (a multi‐centre intervention study) who396
consumednutsaspartofaMeddiethadastatisticallysignificantreduction(p<397
0.05) in LDL‐cholesterol and the LDL/HDL‐cholesterol ratio. 49 The398
cardioprotectiveeffectsofnutsmayberelatedtotheirrelativelyhighproportion399
of unsaturated fatty acids such as MUFA and linoleic acid. 50 The400
hypocholesterolaemiceffectsofnutsmayalso, inpart,berelatedto theirquite401
high phytosterol content. Pistachio nuts ‐ a common aperitif nut in402
Mediterraneancountries‐havethehighestphytosterolcontentofallnuts(279403
mg/100 g), and oil‐ and dry‐roasted peanuts and peanut butter also contain404
moderate levelsofphytosterols (135mg/100g inoil‐roastedpeanuts). 51Tree405
nuts,which constitute a higherproportionof nuts inMediterranean countries,406
haveadditionalnutrientsincluding:ahighvitaminEcontentinalmonds,52high407
levels ofα‐linolenic acid in walnuts,50 and, when eaten with their skins, high408
levelsofphenolicantioxidants.53409
410
In conclusion, although there is good evidence that tree nuts have411
hypocholesterolaemic properties, there is currently no strong evidence for412
hypocholesterolaemic properties for peanuts. Moreover, high levels of salt in413
many peanut products preclude high intake, especially in subjects with high414
blood pressure, and peanuts are also relatively high in saturated fat including415
palmiticacid.416
417
19
Cereals418
It is difficult to appropriately assess the effects of cereals from a nutritional419
questionnaire,sincecerealscanberefinedorwholegrain,saltedorsweetened,420
all factors that strongly influencehealth. InoneMDSwhereneithersweetened421
cereals nor bread from fast‐food were scored for, it was still not possible to422
clearlyevaluatetheeffectsofcereals.11Wholegraincerealsareverylikelytobe423
abettermarkerforhealthbenefits,butthisevaluationcanonlybeappliedwhen424
whole grain cereals are consumed by a high percentage of the subjects under425
study.54Anotherapproachistoscorerefinedcerealsnegatively.55426
427
Legumes428
Legume consumption excluding peas and green beans (ie pulses) showed a429
gradient of consumption in cohorts from the EPIC study,with higher levels of430
consumptioninsoutherncountriesandlowerlevelsinnortherncountries.32As431
well as quantitative differences, the preferred types of legumes in these two432
regions also differs. Commonly consumed legumes inMediterranean countries433
include chickpeas, lentils and fava (broad) beans, although there are national434
differencesegchickpeaconsumptionishighinSpainandintakeoffavabeansis435
highinEgypt.IntheUK,chickpeas,favabeansandlentilsaremainlyconsumed436
byethnicminorities,andthemajortypesof legumes intheUKdietarecanned437
"bakedbeansinsauce"andgardenpeas.33438
439
LegumeconsumptionisassociatedwithadecreasedriskofCHDandCVD,56and440
consumptionofvarious legumes, includingbakedbeans (whicharemade from441
20
haricotbeans),57hasbeenfoundinarecentmeta‐analysistolowercholesterol442
levels. 58 Legumes have excellent nutritional value and were ranked as an443
important source in the US diet for fibre, phytosterols, folate, vitamin B6,444
flavonols, favan‐3‐olsandvariousminerals. 34Both fibreandphytosterolsmay445
belinkedtothehypocholesterolaemiceffectsoflegumes.59InrelationtotheUK446
diet,gardenpeascontain134mg/100gphytosterols,60whichiscomparableto447
amountsfoundinotherpulses,andharicotbeansareagoodsourceoffibre.This448
suggests that legumes commonly consumed in the UK may have some of the449
cardioprotective effects of the pulses more commonly associated with a Med450
diet.29451
452
Fish453
There isgoodevidencethat fishconsumption lowerstheriskofcardiovascular454
mortality. 61 The most important bioactive nutrients in fish are generally455
considered to be the n‐3 LC‐PUFAs EPA and docosahexaenoic acid (DHA). A456
recent systematic review concluded thatmarine n‐3 LC‐PUFAs are effective in457
preventingcardiovascularevents,cardiacdeathandcoronaryevents,especially458
inpersonswithhighcardiovascular risk. 62There is lessevidence fora roleof459
LC‐PUFAsinthepreventionofcancer.63460
461
Levelsofn‐3LC‐PUFAsareconsiderablyhigherinoilyfishthanwhitefish.Since462
there isvariability intherelativeproportionsofoilyandnon‐oily fishbetween463
countries,measuringtotalfishintakemaynotreflecttheintakeofn‐3LC‐PUFAs464
for a givenpopulation.The relative amountsof oily andnon‐oily fishdoesnot465
21
followanorthsouthgradientsinceconsumptionof"veryfattyfish",definedby466
Welchetalasincludingherrings,kippersandmackerel,ishighinScandinavian467
countries. 64Bycontrast, theproportionof these typesof fishconsumed in the468
UK isrelatively low,andso in theUKtotal fishconsumptionwillbeassociated469
withaproportionallylowerintakeofn‐3LC‐PUFAsfromfishthanisthecasein470
someothercountries.471
472
Awide rangeof factors including the fish’s food source influence theLC‐PUFA473
contentofoily fish,andthis isparticularly importantwhenconsideringfarmed474
fish,anincreasinglyimportantdietarysource.Farmedsalmonisamajorsource475
ofn‐3PUFAsintheUKdiet.Salmon,likeothersalt‐waterfish,hasonlyalimited476
capacity to synthesise LC‐PUFAs and instead obtains LC‐PUFAs from its feed.477
Whenfedfishoils,farmedsalmonareanexcellentsourceofLC‐PUFAs,butthere478
are increasing environmental and commercial pressure on fish farmers to use479
non‐marinesourcesofoils.ThispracticecandrasticallyreduceDHAlevels.For480
example, levelsof17gDHA/100gtotal fattyacidsweredeterminedinsalmon481
thatwere fed fishoils,whereas levels of5 gDHA/100g total fatty acidswere482
found in salmon fed plant oils. 65 Feeding fish oils during the last fewmonths483
beforemarketingthefishisonetechniquethatcanrestoreEPAandDHAtofish484
fed a diet that has been predominantly vegetable‐based. 66 Hence, the food485
supplyof farmed fish is an important considerationwhenassessing thehealth486
benefitsofoilyfish.487
488
22
Pan fryingwith olive oil is one of themost popularways of preparing fish in489
Mediterranean countries and fish fried in virgin olive oil has been found to490
absorbsignificantquantitiesofantioxidantphenolics,terpenicacidsandvitamin491
Efromtheoil.67Hence,theremaybeincidentalbenefitsoffryingfishinoliveoil.492
Ontheotherhand,therecanalsobeanexchangeoffattyacidsbetweenthosein493
thefishwiththoseinthefryingoil.N‐3fattyacids insardinesfellbetween2‐3494
foldwhentheywerefriedineithersunfloweroiloroliveoilandtherewasarise495
inn‐6fattyacids.68496
497
Inconclusion, the typeof fish, itsdiet, andhow it ispreparedcontribute to its498
nutritional content, and these factors may vary significantly between499
Mediterraneanandnon‐Mediterraneancountries.500
501
Dairyproduce502
The preferred types of dairy produce consumed in many Mediterranean503
countries are significantly different to those consumed in non‐Mediterranean504
countries. In most Mediterranean countries, Spain being the main exception,505
proportionally less milk and milk beverages and more cheese and yogurt are506
consumedthaninNorthEuropeancountries.69Significantly,a largenumberof507
the cheeses in Mediterranean countries are made from sheep's milk (eg508
Roquefort and tomme from the Pays Basque region of Southern France,509
manchego fromSpain,and feta fromGreece)andgoat'smilk (suchas thewide510
rangeofchevresfromSouthernFrance).Bycontrast,themaindairyproducein511
23
Northern Europe is cow'smilk and cow'smilk cheese such as "cheddar type"512
hardcheeseintheUK.513
514
Althoughcheesesmadefromgoatandsheepmilkhaveasimilartotalsaturated515
fatcontentascheesesmadefromcowmilk,thecompositionofthesaturatedfats516
is different since goat and sheep milk are richer in medium chain fatty acids517
(MCFAs)ie<12carbonatoms(<12C).TheseMCFAsincludecaproicacid(C6:0),518
caprylic(C8:0)andcapric(C10:0)(thenamesarederivedfromtheLatincaper519
foragoat)(Table1).70Forexample, freshgoatcheesewith40%fatcomprises520
15%<12CFAswhereastheseFAsonlyconstitute7%inacomparablecow’smilk521
cheese.Similarly, thefatcontentofRoquefortcheesecomprises15%<12CFAs522
and23%palmiticacid.Bycomparison,acow’smilkfattycheesecomprises33%523
palmitic acid (16C), themost atherogenic SFA. 71 MCFAs are non‐atherogenic,524
andaredirectlyoxidisedintheliverthusreducingtheiraccumulationinadipose525
tissue.SomeepidemiologicalevidencesupportingthebeneficialeffectsofMCFAs526
camefromtheNurses'HealthStudy(aprospectivecohortstudyincludingmore527
than 80,000 US females) which showed that in contrast to LCFAs, intake of528
MCFAswasnotsignificantlyassociatedwiththeriskofCHD.72529
530
The composition of milk is influenced by the animal's diet and this can have531
beneficial effects. Goats and sheep aremore likely than cows to be raised on532
naturalpasture.Pasture isrich inALAandgivesrisetohigher levelsofALAin533
theanimal'slipidswhichcanbedesaturatedtoEPA,ann‐3LC‐PUFAwithanti‐534
inflammatoryproperties.535
24
536
HighconsumptionofdairyproducewasconsideredtobedetrimentalintheMDS537
devisedbyTrichopoulouandcolleagues.Thisisbecausethedairyproducewas538
rarely low‐fat and longer chain SFAs have detrimental effect on cholesterol539
levels.Thereisnowatrendinmanycountriestowardslow‐fatdairyproduce.In540
theUK,weekly purchases ofwholemilk have steadily decreased year on year541
(2655ml/weekin1974downto352ml/weekin2010)withaconcomitantrise542
in the purchase of skimmed milk (mostly semi‐skimmed). 33 Similarly, milk543
consumption in the US is mostly low‐fat. However, not all saturated fats are544
harmful.Myristicacid (14C),present in themilkof ruminants, isnecessary for545
the myristoylation of several functional proteins, 73 and it is not atherogenic546
whentheexogenoussourceis<2%ofthetotalenergyintake.Also,thenatural547
trans‐FAtrans‐palmitoleicacid(cis‐16:1n‐7), levelsofwhichcorrelatestrongly548
with whole fat dairy consumption, was shown to be associated with lower549
metabolicriskfactors.74Theseobservationsquestiontherelevanceofthelow‐550
fatmilkrecommendationsinmanycountries.551
552
Meat553
Arecentmeta‐analysisconcludedthatredmeatconsumptionisassociatedwith554
asmallincreaseintheriskforcolorectalcancer,75althoughotheranalyseshave555
concludedthatbecauseofpossibleconfoundingthecurrentevidenceforsuchan556
associationisweak.76Ithasbeensuggestedthatthepro‐carcinogeniceffectsof557
haem iron, a putative carcinogen in red meat, can be suppressed by various558
dietaryconstituentssuchaschlorophyll,calciumandantioxidantvitamins(Cand559
25
E), 77,78 although the possible relevance of these interactions in relation to the560
Med diet is not known. Marinating, a technique traditionally used in561
Mediterranean cuisine to tenderise cheap cuts, may also potentially have562
beneficialhealtheffects.Cookingmeatathightemperatures,suchasfryingand563
grilling, generates carcinogenic heterocyclic aromatic amines (HA), and564
marinades that contain virgin olive oil, onions, garlic, herbs or redwine have565
highantioxidantcapacityandhavebeenshowntoinhibitHAformation.79‐81566
567
The geographical characteristics of the Mediterranean favour small livestock568
withspecialisedfeedinghabits:sheepandgoatscantakeadvantageofthehilly569
landscapeandofMediterraneangrazing,whereaspigspreferopenwildspaces570
suchasinholm‐oakforest.BecausethepastureofMediterraneananimalsused571
toproducedairyandmeatisricherinPUFAsthanthatoftheequivalentanimal572
givenanimal feeds, theirFAprofile ishealthier.This is especially true forpigs573
running inopenspaces inCorsicaandSardinia‐ theirmeat is leaner,andtheir574
fat consists of 40 to 50%MUFA. Togetherwith the higher content of LA, this575
resultsinlessSFAinthefatcomposition(Table2).71,82576
577
Otherfoods578
Anumber of other foodsnot generally scored for inMDSmayhave important579
healthbenefits foraMeddiet.These includeherbs (also takenasherbal teas),580
wild greens, and pumpkin, sunflower and other types of seeds. Local581
consumption of these can be high, and many are very rich sources of582
26
phytochemicals(forexample,pumpkinseedscontainhighlevelsofphytosterols)583
andothernutrients.2584
585
Notable by their absence in a traditional Med diet aremodern fast foods and586
sugar‐sweeteneddrinks.20Fastfoodscanbeamajorsourceofsaltandtransfats587
‐bothwell‐knownriskfactorsforCVD,andbothfastfoodsandsugar‐sweetened588
drinksarepositivelyassociatedwithlong‐termweightgain.83Anevaluationof589
fast foods and sugar‐sweetened drinks is only rarely included inMDS. In one590
such study from rural Lebanon it was found that when food consumption591
deviated from a traditionalMed diet by including refined cereals and pastries592
and sugar‐sweetened drinks, there was an increase in obesity and visceral593
adiposity.55594
595
Alcohol596
Moderatealcoholconsumption(definedasmenconsuming10gto<50gperday597
andwomenconsuming5gto<25gperday)isassumedtobebeneficialinMDS5.598
This is most strongly linked to a reduced risk of CVD. The most compelling599
mechanism to explain the cardioprotective effects of moderate alcohol600
consumptionistheincreaseinlevelsofHDL‐cholesterol;beneficialeffectsonthe601
vasculaturemayalsobeinvolved.602
603
A number of studies fromnon‐Mediterranean countries have shown that even604
moderatealcoholconsumptionincreasestheriskofcanceratsomesites,suchas605
27
thebreast.84,85Bycontrast,aMeddietreducesoverallcancerrisk,andthereis606
noevidencefromspecificanalysisofthealcoholcomponentthatmoderatelevels607
of consumption in the context of a Med diet increases cancer risk. 86 One608
important factor that might contribute to these disparities between609
Mediterranean and non‐Mediterranean countries with regard to the risk of610
cancerwithmoderatealcohol consumption isdrinkingpattern. 87Whereas the611
custom in Mediterranean countries is to drink in moderation with a meal,612
drinking outside mealtimes and binge drinking are more prevalent in North613
European countries. 88,89 Drinkingwith ameal slows the rate of absorption of614
alcohol from thegutwhereasdrinkingonanempty stomach raisesabsorption615
rates, and may increase blood alcohol concentrations to levels that saturate616
alcohol metabolic pathways resulting in the production of carcinogenic617
metabolites. Dietary folates may also possibly influence the cancer risks of618
drinking. Some studies, 90 but not all, 84 have shown that folates reduce the619
cancer‐associatedeffectsofmoderatealcoholconsumption.Itisnoteworthythat620
theMed diet has particularly high levels of folate consumption, and there is a621
goodcorrelationbetweenfolateconsumptionandadherencetoaMeddiet.91622
623
Alcohol consumption in Mediterranean countries is typically associated with624
wine, mostly red which is drunk with a meal, whereas beer and spirits are625
consumedingreaterquantitiesinsomenon‐Mediterraneancountries.88Thereis626
someevidencethatoverandabovetheeffectsofalcohol,phenolicsthatoccurin627
redwinemayhavespecificcardioprotectiveeffects.Distinguishingbetweenthe628
effects of alcohol and phenols in wine can be undertaken using dealcoholised629
28
wine.Forexample,dealcoholisedredwineretainedtheabilityofcompletewine630
tomodulatedleucocyteadhesionmolecules,importantinflammatorybiomarkers631
relatedtoatherosclerosisinsubjectsathighriskofCVD.92Studiesalsosuggest632
thatdrinkingwinewithamealmayconferadditionalcardioprotectiveeffects.93633
A large number of studies have shown that dealcoholised wine increases634
postprandial total antioxidant capacity and reduces postprandial rises in635
oxidisedLDL‐cholesterol,animportantriskfactorforCVD.46,94,95636
637
In summary, drinking moderate amounts of red wine with a meal may have638
superior health benefits compared to other types of drinking, and this is not639
assessed when measuring consumption of alcohol in MDS. Hence, factors not640
monitoredinMDS‐suchasdrinkingpattern,otherdietaryconstituentsandtype641
of alcohol ‐ are important factors to considerwhenweighing up the risks and642
benefitsassociatedwithmoderatealcoholconsumption.643
644
Discussion645
In the traditional concept of the Med diet, there are various food habits and646
lifestyleaspectsthatmaybeabsentinnon‐Mediterraneanpopulations.Wehave647
describedhowtheaprioriMDSfirstdescribedbyTrichopoulouandcolleagues648
in2003thereforeneededtobeadaptedforspecificpopulations;butineffectitis649
nolongerMediterraneansincethereareanumberofnutritionalcharacteristics650
which are different from the typical Mediterranean diet. Determining if these651
adaptationsinfluencehealthoutcomesisimportantwhenapplyingaMeddietto652
populations in non‐Mediterranean countries. One lifestyle factor that may be653
29
important is physical activity, and in fact this is generally taken into654
consideration in questionnaires. But for others, such as meal structure, the655
organisationofmealsduringtheday,conviviality,andtakingasiesta,itwouldbe656
necessarytoextendthescopeofquestionnairesundertakenaspartofevaluating657
theMDS.Althoughmoretimeconsuming,informationonthesourceofmeatand658
dairy products would also be helpful by assessing the importance of specific659
categories of SFA, and the level of PUFAs. This could be supported with660
biomarkermeasurementsoffattyacidsandothernutrients.11Anotherbiological661
measurementofinterestisvitaminDstatus.662
663
It became evident during the early development of MDS that the first MDS,664
originallydevelopedforaMediterraneanpopulation,wouldhavetobeadapted665
for non‐Mediterranean populations in order to take into consideration the666
genuineeatinghabitsofnon‐Mediterraneanpopulations.19,54,55Infact,takinga667
nutritional survey aposteriori in order to identify dietary patterns is the best668
way to reveal the healthiest type of diet among the eating habits of this669
population.ThesepatternscouldthenbecomparedtotheoriginalMeddiet,and670
differencesbetween the twopatterns canbe identified.Using this approach to671
adapting theMDS, some important featureshavebeen identified. For example,672
therMed‐score19andthenewaposterioriMDS17 identified the importanceof673
taking olive oil consumption into account, and for the latter the necessity of674
consideringanabsoluteamountoffoodinrelationtoitsbeneficialeffect.675
676
30
Measuringplasmalevelsofselectednutrientsthatarespecifictoauniquesource677
orclassof food(e.g.LCn‐3PUFAand fish)mayofferamoreprecisewaythan678
FoodFrequencyQuestionnairesofassessingtheroleofspecificnutrientsforthe679
healthbenefitsofthetraditionalMeddiet.WhenapplyingtheMeddiet innon‐680
Mediterraneanpopulations,particularlyrelevantnutrientsarethoselinkedwith681
healthwhose levels are likely tobe influencedby culinaryand lifestyle factors682
that vary significantly between Mediterranean and non‐Mediterranean683
populations. One candidate nutrient is folate, since dietary intake of folate is684
associated with a reduced risk of a range of chronic disorders such as colon685
cancer. 96Folate levels in foodsare influencedbyculinarypractices: it iswater686
soluble and heat labile and boiling results in leaching of folate into cooking687
water.41Moreover,folatelevelsvarywidelybetweenvegetablessinceitoccurs688
mainly in green leafy vegetables and so will not be specifically assessed by689
quantifying total vegetable intake. Hence, folate plasma levels may not690
necessarilybereflectedinMDS.691
692
SufficientplasmalevelsofothernutrientsthatarelinkedtoadherencetoaMed693
dietmayalsobe important inorder fornon‐Mediterraneans toachieve the full694
benefits of a traditional Med diet. Gerber et al. found a significant correlation695
betweenplasmalevelsofα‐tocopherol,β‐carotene,EPAandDHAwiththeMed696
dietqualityindex,especiallywhenthesenutrientswerecombinedinacomposite697
index, thus establishing that plasma values of these nutrients correlated with698
goodadherencetoaMeddiet.11SubjectsinNorthernSpainwithhigherMeddiet699
adherence,asmeasuredbytwodietaryindexes,hadsignificantlyhigherplasma700
31
concentrations of β‐carotene, folates, vitamin C, α‐tocopherol and HDL701
cholesterol. 91 Carotenoids may also be useful biomarkers because of their702
associationwithriskofcanceratcertainsitesandbecausefoodpreparationcan703
influence bioavailability. However, correlating plasma concentrations of a704
particular nutrient with disease reduction is not straightforward since food705
plantscontainmanynutrients thatareconsumedat thesametimee.g.mostof706
thefruitandvegetablescontainingcarotenoidscontainalsolignans.707
708
Conclusions709
Although there is extensive epidemiological evidence supporting the health710
benefitsofaMediterraneandiet,thisismostlyderivedfromMediterraneandiet711
scoresthatdonotfullytakeintoconsiderationthemanypotentiallyconfounding712
factors between Mediterranean and non‐Mediterranean populations. We have713
identifiedanumberof factors thatmayconfound theassessmentof thehealth714
benefitsofaMeddietinMediterraneancountries,bothbecauseofthenatureof715
the food itself and also due to aspects of food production and preparation.716
Lifestylefactorssuchasmealpatternsandexposuretosunlightmayalsoactas717
confounders. Improving Mediterranean diet scores and measuring plasma718
nutrient levelsmayhelpmitigate the effects of these confounders. Taking into719
considerationtheseconfoundersmayhave importanthealth implicationswhen720
implementingaMediterraneandietinnon‐Mediterraneanpopulations.721
722
Acknowledgements723
Theauthorshavenorelevantintereststodeclare724
32
725
References726
1. GerberM. Santé etAlimentationMéditerranéenne auQuotidien. Aix‐en‐727
Provence:Édisud;2004.728
2. HoffmanR,GerberM.TheMediterraneandiet:health&science.Oxford:729
Wiley‐Blackwell;2012.730
3. KeysA,MenottiA,KarvonenMJ,etal.Thedietand15‐yeardeathratein731
thesevencountriesstudy.AmJEpidemiol.1986;124(6):903‐915.732
4. Scali J, Richard A, Gerber M. Diet profiles in a population sample from733
MediterraneansouthernFrance.PublicHealthNutr.2001;4(2):173‐182.734
5. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a735
Mediterranean diet and survival in a Greek population. N Engl J Med.736
2003;348(26):2599‐2608.737
6. McKeownPP, LoganK,McKinleyMC,Young IS,Woodside JV. Session4:738
CVD, diabetes and cancer: Evidence for the use of the Mediterranean diet in739
patientswithCHD.ProcNutrSoc.2010;69(1):45‐60.740
7. Logan KJ, Woodside JV, Young IS, McKinley MC, Perkins‐Porras L,741
McKeown PP. Adoption and maintenance of a Mediterranean diet in patients742
with coronary heart disease from a Northern European population: a pilot743
randomisedtrialofdifferentmethodsofdeliveringMediterraneandietadvice.J744
HumNutrDiet.2010;23(1):30‐37.745
8. GerberM.Thecomprehensiveapproachtodiet:acriticalreview. JNutr.746
2001;131(11Suppl):3051S‐3055S.747
33
9. Hu FB. Dietary pattern analysis: a new direction in nutritional748
epidemiology.CurrOpinLipidol.2002;13(1):3‐9.749
10. WillettWC,SacksF,TrichopoulouA,etal.Mediterraneandietpyramid:a750
cultural model for healthy eating. Am J Clin Nutr. 1995;61(6 Suppl):1402S‐751
1406S.752
11. GerberMJ, Scali JD,MichaudA, et al. Profiles of a healthful diet and its753
relationshiptobiomarkersinapopulationsamplefromMediterraneansouthern754
France.JAmDietAssoc.2000;100(10):1164‐1171.755
12. Alberti‐Fidanza A, Fidanza F. Mediterranean Adequacy Index of Italian756
diets.PublicHealthNutr.2004;7(7):937‐941.757
13. Trichopoulou A, Orfanos P, Norat T, et al. Modified Mediterranean diet758
andsurvival:EPIC‐elderlyprospectivecohortstudy.Bmj.2005;330(7498):991‐759
998.760
14. JakobsenMU, O'Reilly EJ, HeitmannBL, et al.Major types of dietary fat761
andriskofcoronaryheartdisease:apooledanalysisof11cohortstudies.AmJ762
ClinNutr.2009;89(5):1425‐1432.763
15. RichardsonS,GerberM,CeneeS.Theroleoffat,animalproteinandsome764
vitaminconsumptioninbreastcancer:acasecontrolstudyinsouthernFrance.765
IntJCancer.1991;48(1):1‐9.766
16. GerberM, Richardson S. Re: Consumption of olive oil and specific food767
groups in relation to breast cancer risk in Greece. J Natl Cancer Inst.768
1995;87(13):1021‐1022.769
34
17. Sofi F, Abbate R, Gensini GF, Casini A, Trichopoulou A, Bamia C.770
Identification of change‐points in the relationship between food groups in the771
Mediterraneandietandoverallmortality:an'aposteriori'approach.EurJNutr.772
2012;51(2):167‐172.773
18. FungTT,HuFB,McCulloughML,NewbyPK,WillettWC,HolmesMD.Diet774
qualityisassociatedwiththeriskofestrogenreceptor‐negativebreastcancerin775
postmenopausalwomen.JNutr.2006;136(2):466‐472.776
19. BucklandG, AgudoA, Lujan L, et al. Adherence to aMediterraneandiet777
and risk of gastric adenocarcinoma within the European Prospective778
Investigation into Cancer and Nutrition (EPIC) cohort study. Am J Clin Nutr.779
2010;91(2):381‐390.780
20. Tessier S, Gerber M. Comparison between Sardinia and Malta: the781
Mediterraneandietrevisited.Appetite.2005;45(2):121‐126.782
21. NaskaA,OikonomouE,TrichopoulouA,PsaltopoulouT,TrichopoulosD.783
Siesta inhealthyadultsandcoronarymortality in thegeneralpopulation.Arch784
InternMed.2007;167(3):296‐301.785
22. CheungFS, LovicuFJ,Reichardt JK.Currentprogress inusingvitaminD786
anditsanalogsforcancerpreventionandtreatment.ExpertRevAnticancerTher.787
2012;12(6):811‐837.788
23. Danik JS, Manson JE. Vitamin D and cardiovascular disease. Curr Treat789
OptionsCardiovascMed.2012;14(4):414‐424.790
24. Warleta F, Campos M, Allouche Y, et al. Squalene protects against791
oxidativeDNAdamage inMCF10Ahumanmammary epithelial cells but not in792
35
MCF7 and MDA‐MB‐231 human breast cancer cells. Food Chem Toxicol.793
2010;48(4):1092‐1100.794
25. Buck K, Zaineddin AK, Vrieling A, Linseisen J, Chang‐Claude J. Meta‐795
analysesoflignansandenterolignansinrelationtobreastcancerrisk.AmJClin796
Nutr.2010;92(1):141‐153.797
26. HoffmanR,GerberM.Virginoliveoilasasourceofphytoestrogens.EurJ798
ClinNutr.2012;66(10):1180.799
27. Granados‐Principal S, Quiles JL, Ramirez‐Tortosa CL, Sanchez‐Rovira P,800
Ramirez‐TortosaMC. Hydroxytyrosol: from laboratory investigations to future801
clinicaltrials.NutrRev.2010;68(4):191‐206.802
28. Gomez‐Alonso S, Fregapane G, Salvador MD, Gordon MH. Changes in803
phenoliccompositionandantioxidantactivityofvirginoliveoilduring frying. J804
AgricFoodChem.2003;51(3):667‐672.805
29. Holdsworth M, Gerber M, Haslam C, et al. A comparison of dietary806
behaviourinCentralEnglandandaFrenchMediterraneanregion.EurJClinNutr.807
2000;54(7):530‐539.808
30. VanDiepenS,ScholtenAM,KorobiliC,etal.GreaterMediterraneandiet809
adherenceisobservedinDutchcomparedwithGreekuniversitystudents.Nutr810
MetabCardiovascDis.2011;21(7):534‐540.811
31. Kafatos A, Verhagen H, Moschandreas J, Apostolaki I, Van Westerop JJ.812
Mediterranean diet of Crete: foods and nutrient content. J Am Diet Assoc.813
2000;100(12):1487‐1493.814
36
32. AgudoA,SlimaniN,OckeMC,etal.Consumptionofvegetables, fruitand815
other plant foods in the European Prospective Investigation into Cancer and816
Nutrition (EPIC) cohorts from 10 European countries. Public Health Nutr.817
2002;5(6B):1179‐1196.818
33. DEFRA. UK household purchased quantities of food and drink.819
http://www.defra.gov.uk/statistics/foodfarm/food/familyfood/datasets/.2010.820
34. Pennington JAT, Fisher RA. Food component profiles for fruit and821
vegetable subgroups. Journal of Food Composition and Analysis. 2010;23:411‐822
418.823
35. WCRF/AICR. World Cancer Research Fund / American Institute for824
CancerResearch.Food,Nutrition,PhysicalActivity,andthePreventionofCancer:825
aGlobalPerspective.WashingtonDC:AICR;2007.826
36. DuPont MS, Mondin Z, Williamson G, Price KR. Effect of variety,827
processing, and storage on the flavonoid glycoside content and compositionof828
lettuceandendive.JAgricFoodChem.2000;48(9):3957‐3964.829
37. Hollman PC, Geelen A, Kromhout D. Dietary flavonol intake may lower830
strokeriskinmenandwomen.JNutr.2010;140(3):600‐604.831
38. Buchner FL, Bueno‐de‐MesquitaHB, RosMM, et al. Variety in Fruit and832
VegetableConsumptionandtheRiskofLungCancerintheEuropeanProspective833
Investigation into Cancer and Nutrition. Cancer Epidemiol Biomarkers Prev.834
2010;19(9):2278‐2286.835
37
39. GoltzSR,CampbellWW,ChitchumroonchokchaiC,FaillaML,FerruzziMG.836
Mealtriacylglycerolprofilemodulatespostprandialabsorptionofcarotenoidsin837
humans.MolNutrFoodRes.2012;56(6):866‐877.838
40. vanHetHofKH,WestCE,WeststrateJA,HautvastJG.Dietaryfactorsthat839
affectthebioavailabilityofcarotenoids.JNutr.2000;130(3):503‐506.840
41. Delchiera N, Reicha M, Renard CMGC. Impact of cooking methods on841
folates,ascorbicacidandluteiningreenbeans(Phaseolusvulgaris)andspinach842
(Spinaceaoleracea).FoodScienceandTechnology.2012;49(2):197‐201.843
42. Song L, Thornalley PJ. Effect of storage, processing and cooking on844
glucosinolate content of Brassica vegetables. Food Chem Toxicol.845
2007;45(2):216‐224.846
43. Miglio C, Chiavaro E, Visconti A, Fogliano V, Pellegrini N. Effects of847
differentcookingmethodsonnutritionalandphysicochemicalcharacteristicsof848
selectedvegetables.JAgricFoodChem.2008;56(1):139‐147.849
44. Jansen M, Hectors K, O’Brien N, Guisez Y, Potters G. Plant stress and850
humanhealth:DohumanconsumersbenefitfromUV‐Bacclimatedcrops?Plant851
Science.2008;175:449‐458.852
45. ChongMF,MacdonaldR,LovegroveJA.FruitpolyphenolsandCVDrisk:a853
reviewofhumaninterventionstudies.BrJNutr.2010;104Suppl3:S28‐39.854
46. Burton‐Freeman B. Postprandial metabolic events and fruit‐derived855
phenolics:areviewofthescience.BrJNutr.2010;104Suppl3:S1‐14.856
47. JenabM,SabateJ,SlimaniN,etal.Consumptionandportionsizesoftree857
nuts, peanuts and seeds in theEuropeanProspective Investigation intoCancer858
38
and Nutrition (EPIC) cohorts from 10 European countries. Br J Nutr. 2006;96859
Suppl2:S12‐23.860
48. SabateJ,OdaK,RosE.Nutconsumptionandbloodlipid levels:apooled861
analysisof25interventiontrials.ArchInternMed.2010;170(9):821‐827.862
49. EscurriolV,CofanM,SerraM,etal.Serumsterolresponsestoincreasing863
plant sterol intake from natural foods in the Mediterranean diet. Eur J Nutr.864
2009;48(6):373‐382.865
50. Ros E, Mataix J. Fatty acid composition of nuts‐‐implications for866
cardiovascularhealth.BrJNutr.2006;96Suppl2:S29‐35.867
51. PhillipsKM,RuggioDM,Ashraf‐KhorassaniM.Phytosterolcompositionof868
nuts and seeds commonly consumed in the United States. J Agric Food Chem.869
2005;53(24):9436‐9445.870
52. Lopez‐Uriarte P, Bullo M, Casas‐Agustench P, Babio N, Salas‐Salvado J.871
Nutsandoxidation:asystematicreview.NutrRev.2009;67(9):497‐508.872
53. Blomhoff R, CarlsenMH, Andersen LF, Jacobs DR, Jr. Health benefits of873
nuts:potentialroleofantioxidants.BrJNutr.2006;96Suppl2:S52‐60.874
54. Fung TT, Hu FB, Holmes MD, et al. Dietary patterns and the risk of875
postmenopausalbreastcancer.IntJCancer.2005;116(1):116‐121.876
55. Issa C, Darmon N, Salameh P, Maillot M, Batal M, Lairon D. A877
Mediterraneandietpatternwith lowconsumptionof liquidsweetsandrefined878
cerealsisnegativelyassociatedwithadiposityinadultsfromruralLebanon.IntJ879
Obes(Lond).2011;35(2):251‐258.880
39
56. Bazzano LA, He J, Ogden LG, et al. Legume consumption and risk of881
coronaryheartdiseaseinUSmenandwomen:NHANESIEpidemiologicFollow‐882
upStudy.ArchInternMed.2001;161(21):2573‐2578.883
57. WinhamDM,HutchinsAM,JohnstonCS.Pintobeanconsumptionreduces884
biomarkersforheartdiseaserisk.JAmCollNutr.2007;26(3):243‐249.885
58. BazzanoLA,ThompsonAM,TeesMT,NguyenCH,WinhamDM.Non‐soy886
legume consumption lowers cholesterol levels: ameta‐analysis of randomized887
controlledtrials.NutrMetabCardiovascDis.2011;21(2):94‐103.888
59. Campos‐VegaR,Loarca‐PiñaG,OomahBD.Minor componentsofpulses889
and their potential impact on human health. Food Research International.890
2010;43:461–482.891
60. Piironen V, Toivo J, Puupponen‐Pimia R, Lampi AM. Plant sterols in892
vegetables,fruitsandberries.JSciFoodAgric.2003;83:330–337.893
61. He K. Fish, long‐chain omega‐3 polyunsaturated fatty acids and894
preventionofcardiovasculardisease‐‐eat fishortakefishoilsupplement?Prog895
CardiovascDis.2009;52(2):95‐114.896
62. Delgado‐ListaJ,Perez‐MartinezP,Lopez‐MirandaJ,Perez‐JimenezF.Long897
chainomega‐3 fattyacidsandcardiovasculardisease:asystematicreview.Br J898
Nutr.2012;107Suppl2:S201‐213.899
63. GerberM.Omega‐3fattyacidsandcancers:asystematicupdatereviewof900
epidemiologicalstudies.BrJNutr.2012;107Suppl2:S228‐239.901
40
64. WelchAA,LundE,AmianoP,etal.Variabilityoffishconsumptionwithin902
the 10 European countries participating in the European Investigation into903
CancerandNutrition(EPIC)study.PublicHealthNutr.2002;5(6B):1273‐1285.904
65. Nichols PD, Petrie J, Singh S. Long‐chain omega‐3 oils‐an update on905
sustainablesources.Nutrients.2010;2(6):572‐585.906
66. BellJG,TocherDR,HendersonRJ,DickJR,CramptonVO.Alteredfattyacid907
compositions in atlantic salmon (Salmo salar) fed diets containing linseed and908
rapeseedoils canbepartially restoredbya subsequent fishoil finishingdiet. J909
Nutr.2003;133(9):2793‐2801.910
67. Kalogeropoulos N, Chiou A, Mylona A, Ioannou M, Andrikopoulos N.911
Recovery and distribution of natural antioxidants (a‐tocopherol, polyphenols912
and terpenicacids)afterpan‐fryingofMediterranean finfish invirginoliveoil.913
FoodChemistry.2007;100:509–517.914
68. Sanchez‐Muniz F, Viejo J,MedinaR.Deep‐frying of sardines in different915
culinaryfats.Changesinthefattyacidcompositionofsardinesandfryingfats..J916
AgricFoodChem.1992;40:2252–2256.917
69. HjartakerA,LagiouA,SlimaniN,etal.Consumptionofdairyproductsin918
theEuropeanProspectiveInvestigationintoCancerandNutrition(EPIC)cohort:919
datafrom3595524‐hourdietaryrecallsin10Europeancountries.PublicHealth920
Nutr.2002;5(6B):1259‐1271.921
70. Park YW, Juarez M, Ramos M, Haenlein GFW. Physico‐chemical922
characteristics of goat and sheepmilk. Small Ruminant Research. 2007;68:88‐923
113.924
41
71. McCance RA, Widdowson EM, Paul AA, Southgate DAT. McCance and925
Widdowson'sThecompositionof foods.4threvisedandextendeded./byA.A.926
PaulandD.A.T.Southgate/[forthe]MinistryofAgriculture,FisheriesandFood927
[andthe]MedicalResearchCouncil.ed.London:H.M.S.O.;Amsterdam;Oxford:928
Elsevier/North‐HollandBiomedicalPress;1978.929
72. HuFB,StampferMJ,MansonJE,etal.Dietarysaturatedfatsandtheirfood930
sources in relation to the risk of coronary heart disease in women. Am J Clin931
Nutr.1999;70(6):1001‐1008.932
73. RiouxV,GalatA,JanG,VinciF,D'AndreaS,LegrandP.Exogenousmyristic933
acid acylates proteins in cultured rat hepatocytes. J Nutr Biochem.934
2002;13(2):66‐74.935
74. MozaffarianD,CaoH,KingIB,etal.Trans‐palmitoleicacid,metabolicrisk936
factors, andnew‐onsetdiabetes inU.S. adults: a cohort study.Ann InternMed.937
2010;153(12):790‐799.938
75. Chan DS, Lau R, Aune D, et al. Red and processed meat and colorectal939
cancer incidence: meta‐analysis of prospective studies. PLoS One.940
2011;6(6):e20456.941
76. Alexander DD, Cushing CA. Red meat and colorectal cancer: a critical942
summaryofprospectiveepidemiologicstudies.ObesRev.2011;12(5):e472‐493.943
77. BalderHF,VogelJ,JansenMC,etal.Hemeandchlorophyllintakeandrisk944
of colorectal cancer in the Netherlands cohort study. Cancer Epidemiol945
BiomarkersPrev.2006;15(4):717‐725.946
42
78. Bastide NM, Pierre FH, Corpet DE. Heme iron from meat and risk of947
colorectal cancer: a meta‐analysis and a review of the mechanisms involved.948
CancerPrevRes(Phila).2011;4(2):177‐184.949
79. Persson E, Graziani G, Ferracane R, Fogliano V, Skog K. Influence of950
antioxidants invirginoliveoilon the formationofheterocyclic amines in fried951
beefburgers.FoodChemToxicol.2003;41(11):1587‐1597.952
80. GibisM.Effectofoilmarinadeswithgarlic,onion,andlemonjuiceonthe953
formation of heterocyclic aromatic amines in fried beef patties. J Agric Food954
Chem.2007;55(25):10240‐10247.955
81. BusquetsR,PuignouL,GalceranMT,SkogK.Effectofredwinemarinades956
on the formation of heterocyclic amines in fried chicken breast. J Agric Food957
Chem.2006;54(21):8376‐8384.958
82. Favier JC, Ireland‐Ripert J, Toque C, Feinberg M. TEC‐DOC. Paris:959
Lavoisier;1995.960
83. MozaffarianD,HaoT,RimmEB,WillettWC,HuFB.Changes indietand961
lifestyle and long‐term weight gain in women and men. N Engl J Med.962
2011;364(25):2392‐2404.963
84. AllenNE,BeralV,CasabonneD,etal.Moderatealcoholintakeandcancer964
incidenceinwomen.JNatlCancerInst.2009;101(5):296‐305.965
85. RichardsonS,deVincenziI,PujolH,GerberM.Alcoholconsumptionina966
case‐control study of breast cancer in southern France. Int J Cancer.967
1989;44(1):84‐89.968
43
86. Benetou V, Trichopoulou A, Orfanos P, et al. Conformity to traditional969
Mediterranean diet and cancer incidence: the Greek EPIC cohort. Br J Cancer.970
2008;99(1):191‐195.971
87. BessaoudF,DauresJP.Patternsofalcohol(especiallywine)consumption972
and breast cancer risk: a case‐control study among a population in Southern973
France.AnnEpidemiol.2008;18(6):467‐475.974
88. Sieri S, Agudo A, Kesse E, et al. Patterns of alcohol consumption in 10975
EuropeancountriesparticipatingintheEuropeanProspectiveInvestigationinto976
CancerandNutrition(EPIC)project.PublicHealthNutr.2002;5(6B):1287‐1296.977
89. AndersonP,BaumbergB.AlcoholinEurope.London:InstituteofAlcohol978
Studies.2006.979
90. BagliettoL,EnglishDR,GertigDM,HopperJL,GilesGG.Doesdietaryfolate980
intakemodifyeffectofalcoholconsumptiononbreastcancerrisk?Prospective981
cohortstudy.Bmj.2005;331(7520):807.982
91. Bach‐Faig A, Geleva D, Carrasco JL, Ribas‐Barba L, Serra‐Majem L.983
Evaluating associations between Mediterranean diet adherence indexes and984
biomarkersofdietanddisease.PublicHealthNutr.2006;9(8A):1110‐1117.985
92. Chiva‐Blanch G, Urpi‐Sarda M, Llorach R, et al. Differential effects of986
polyphenols and alcohol of redwine on the expression of adhesionmolecules987
and inflammatory cytokines related to atherosclerosis: a randomized clinical988
trial.AmJClinNutr.2012;95(2):326‐334.989
93. Rehm J, Sempos CT, Trevisan M. Alcohol and cardiovascular disease‐‐990
more than one paradox to consider. Average volume of alcohol consumption,991
44
patternsofdrinking and riskof coronaryheartdisease‐‐a review. J Cardiovasc992
Risk.2003;10(1):15‐20.993
94. NatellaF,MaconeA,RambertiA,etal.Redwinepreventsthepostprandial994
increase in plasma cholesterol oxidation products: a pilot study. Br J Nutr.995
2011:1‐6.996
95. CovasMI,GambertP,FitoM,delaTorreR.Wineandoxidativestress:up‐997
to‐date evidence of the effects of moderate wine consumption on oxidative998
damageinhumans.Atherosclerosis.2010;208(2):297‐304.999
96. Gibson TM, Weinstein SJ, Pfeiffer RM, et al. Pre‐ and postfortification1000
intakeoffolateandriskofcolorectalcancerinalargeprospectivecohortstudy1001
intheUnitedStates.AmJClinNutr.2011;94(4):1053‐1062.1002
1003Table1Mediumchainfattyacidcontentofsheep,goatandcowmilk701004
Fattyacid Fattyacidcomposition(%total)
Sheep Goat Cow
CaproicC6:0 2.9 2.4 1.6
CaprylicC8:0 2.6 2.7 1.3
CapricC10:0 7.8 10.0 3.0
Total 13.3 15.1 5.9
1005
1006
Table2Fatcompositionofvariousmeats(basedon71,82)1007
45
Meat Fattyacids(%totalfat)
Saturatedfattyacids
Monounsaturatedfattyacids
Totalpolyunsaturatedfattyacids
Linoleicacid
Alpha‐linolenicacid
Lamb 52.1 40.5 5.8 5.0 0.8
Pork 43.2 47.6 9.2 8.6 0.6
Beef 56.4 40.3 3.2 2.5 0.7
1008
1009
1010