<b>Bijsluiter</b>. De hyperlink naar het originele document werkt niet meer. Daarom laat Woogle de tekst zien die in dat document stond. Deze tekst kan vreemde foutieve woorden of zinnen bevatten en de opmaak kan verdwenen of veranderd zijn. Dit komt door het zwartlakken van vertrouwelijke informatie of doordat de tekst niet digitaal beschikbaar was en dus ingescand en vervolgens via OCR weer ingelezen is. Voor het originele document, neem contact op met de Woo-contactpersoon van het bestuursorgaan.<br><br>====================================================================== Pagina 1 ======================================================================

<pre>Ingezonden commentaren op het openbare concept van het
achtergronddocument Alcoholhoudende dranken
De volgende organisaties hebben commentaar ingestuurd:
    Federatie Nederlandse Levensmiddelen Industrie
    Kennisinstituut Bier
    Nederlands Instituut voor Alcoholbeleid STAP
    Rijksinstituut voor Volksgezondheid en Milieu
    STIVA Stichting Verantwoorde Alcoholconsumptie
    Trimbos instituut
    Wereld Kanker Onderzoek Fonds
</pre>

====================================================================== Einde pagina 1 =================================================================

<br><br>====================================================================== Pagina 2 ======================================================================

<pre>                                                                                           823985
Van: Christine Grit
Verzonden: maandag 24 augustus 2015 15:55
Aan: GR_RGV2O15
Onderwerp: EGV-015 020 A Respons op vijfde serie achtergronddocumenten Gezondheidsraad RGV
2015 definitief
Geachte mevrouw/heer
Bijgaand doe ik u de respons namens de FNLI toekomen in reactie op de vijfde reeks
achtergronddocumenten van de Gezondheidsraad ten behoeve van het opstellen van de
Richtlijnen goede voeding 2015.
Wij hopen dat u de in de respons opgenomen informatie kunt gebruiken.
Met vriendelijke groet,
Christine Grit
Manager Voeding & Gezondheid
FN LI
fnli.nl 1 voedincivooruit.nl 1 duurzamereten.nl 1 Twitter 1 Linkedln
</pre>

====================================================================== Einde pagina 2 =================================================================

<br><br>====================================================================== Pagina 3 ======================================================================

<pre>                        EGV-015 020 A Cousnitatie respons vijfde ronde achtergronddocunienten Gezondheidsraad         -
                        definitief
                        EGV       15  019      A
                        Notitie
                        Consultatierespons op 5 achtergronddocumenten
                        Onderwerp                Achtergronddocumenten (1) Alcoholhoudende dranken, (2) Eiwit,
                                                 (3) Kalium, (4) Transvetzuren en (5) Visvetzuren.
                        Datum             1 24 augustus 2015
                        Inleiding
                        Als eerste willen we ook bij deze vijfde reeks achtergronddocumenten de Commissie
                        bedanken voor het kunnen inzien van de Werkwijze en de achtergronddocumenten voor
                        de Richtlijnen goede voeding (Rgv) 2015. Ook bij deze set documenten willen we graag
                        de Con-unissie complimenteren met het vele werk dat hiertoe moet zijn uitgevoerd.
                        Het blijft voor ons lastig dat naarmate er meer documenten komen, het steeds
                        onduidelijker wordt om overzicht te houden op de dwarsverbanden tussen
                        voedingsstoffen, voedingssupplementen, voedingsmiddelen en voedingspatronen. Vaak
                        duiken onderwerpen die (deels) al in een bepaald achtergronddocument zijn besproken
                        ook op andere plaatsen op, soms wordt verwezen naar documenten en soms worden in
                        het ene achtergronddocument andere getallen gehanteerd dan voor het andere terwijl
                        deze gelijk zouden kunnen en moeten zijn.
                        Een ander punt dat ons zorgen blijft baren, is dat de keuze voor de top 10 van ziekten er
                        toe bij kan dragen dat bepaalde voedingsgerelateerde aandoeningen niet of slechts heel
                        beperkt zullen worden meegewogen bij het opstellen van de Richtlijnen. Terwijl hier
                        sprake is van aandoeningen die weliswaar niet in de top 10 voorkomen maar wel degelijk
                        grote gevolgen kunnen hebben voor de volksgezondheid en ook voor een deel kunnen
                        worden voorkomen.
                        Terugkomend op de voedingsstoffen/levensmiddelen die in achtergronddocumenten
                        überhaupt aan de orde komen (waarop wij dieper ingaan in onze respons op de eerste
                        reeks achtergronddocumenten), is het onzes inziens een gemis dat er geen aandacht is
                        voor plantensterolen en producten met toegevoegde plantensterolen. Zeker bij een
                        exercitie waarin de preventie van hart- en vaatziekten en waarin het niveau van het LDL
                        cholesterol gehalte als een belangrijke intermediair is meegenomen, valt het op dat er
                        geen aandacht voor is.
             F N II                                                             Respons consultatie vijfde ronde achtergronddocumenten 1 1
FEDERATIE NEDERLANDSE
1[VEN5F.IIDOESEN INDIJSTRI[
</pre>

====================================================================== Einde pagina 3 =================================================================

<br><br>====================================================================== Pagina 4 ======================================================================

<pre>                      EGV-015 020 A Consultatie respons vijfde ronde aclitergronddocunienten Gezondheidsraad        -
                      definitief
                      Los van deze algemene aandachtspunten die ons enige zorgen baren, maken we opnieuw
                      graag van de gelegenheid gebruik om te reageren op de verschillende
                      achtergronddocumenten die bij deze vijfde ronde zijn verspreid voor consultatie. Alle 5
                      de achtergronddocumenten zijn in onze achterban doorgenomen waarbij uiteraard de
                      door de Commissie gestelde vragen zoveel mogelijk centraal hebben gestaan. De reacties
                      op de verschillende documenten volgen vanaf pagina 3 van deze consultatierespons. De
                      documenten worden in alfabetische volgorde behandeld, te beginnen hij
                      ‘Alcoholhoudende dranken’ en eindigend bij ‘Visvetzuren (Eicosapentaeenzuur en
                      docosahexaeenzuur)’.
                      Voor de goede orde zij nog opgemerkt dat de aandachtspunten over de werkwijze die wij
                      in de respons op de eerste reeks achtergronddocumenten hebben weergegeven, ook op
                      deze reeks achtergronddocumenten van toepassing blijven.
            F I4 II                                                           Respons consultatie vijfde ronde achtergronddocumenten 1 2
FEDERATIE NEDERLANDSE
LEVENSMIDDELEN INDUSTRIE
</pre>

====================================================================== Einde pagina 4 =================================================================

<br><br>====================================================================== Pagina 5 ======================================================================

<pre>                         EG V-015 020 A Consultatie respons vijfde ronde achtergronddocu uien ten Gezondlieidsraad      -
                         definitief
                         Alcoholhoudende dranken
                         Opinerkin gen
                         Het is ons niet duidelijk geworden wat nu precies sterke drank is. Zijn dat alle dranken
                         met een bepaald minimum percentage aan alcohol (en welk percentage geldt dan)? Is
                         daarbij gecorrigeerd voor de aanwezigheid van andere voedingsstoffen (zoals
                        bijvoorbeeld suikers ix likeurdranken)? Zijn in alle studies de definities hetzelfde?
                         Het is duidelijk bij de innamecijfers dat cider is meegeteld bij hier. Is de voedingswaarde
                         echter gelijk afgezien van het alcoholgehalte? Is er rekening mee gehouden dat sommige
                        bieren hogere gehaltes aan alcohol hebben dan andere? We vragen ons bovendien af in
                        hoeverre het meetellen van versterkte wijnen zoals port en sherry bij” wijn” de resultaten
                        niet zullen vertekenen. Daarbij komt dan nog dat onduidelijk is of in de studies dezelfde
                        dranken steeds zijn meegeteld als is weergegeven in de tabel met wat in Nederland
                        wordt gedronken.
                        Als laatste vragen we ons af in hoeverre bepaalde mixdrankjes zijn meegenomen in het
                        achtergronddocument. Qua hoeveelheid alcohol per 100 gram bevinden deze zich dichter
                        in de buurt van de wijnen dan de sterke drank hoewel ze vaak met sterke drank worden
                        gemaakt.
            F N II                                                               Respons consultatie vijfde ronde achtergronddocumenten 1 3
FEDERATIE NEDERLANDSE
LSVENSMIDDESTT4 IT1DUSTRIE
</pre>

====================================================================== Einde pagina 5 =================================================================

<br><br>====================================================================== Pagina 6 ======================================================================

<pre>Van: Ivonne Sleutels
Verzonden: vrijdag 21 augustus 2015 11:01
Aan: Javanmardi, M. (Mitra)
CC: Schoten, E.J. (Eert); Allers, J.M.; GR_Webmaster; Aafje Sierksma
Onderwerp: Re: Vijfde comnientaarroncle Richtlijnen goede voeding 2015
Geachte heer, mevrouw.
In de bijlage vindt u het commentaar van dr. ir. Aafje Sierksma, directeur Kennisiristituut Bier
op het achtergronddocument over alcoholhoudende dranken.
Zou u kunnen bevestigen dat u dit document in goed orde hebt ontvangen?
Met vriendelijke groet.
Ir. Ivonne Sleutels
Conirnunicatiemedewerker
Kennisinstituut Bier
Postbus 590 1 6700 AN 1 Wageningen
Website     Facebook Twitter
</pre>

====================================================================== Einde pagina 6 =================================================================

<br><br>====================================================================== Pagina 7 ======================================================================

<pre></pre>

====================================================================== Einde pagina 7 =================================================================

<br><br>====================================================================== Pagina 8 ======================================================================

<pre>Aan de Gezondheidsraad
                                                    kennis  k
                                                  instituut uler
Wageningen, 21 augustus 2015
Geachte Gezondheidsraad,
Graag stuur ik u ons commentaar op uw document
‘Concept —Achtergronddocument Richtlijnen goede
voeding2Ol5 Alcoholhoudende dranken’.
                —
Allereerst wil ik aangeven dat het eerdere document:
‘Concept —Achtergronddocument Richtlijnen goede
voeding 2015 Alcohol’ over het algemeen overeenkomt
                —
met onze interpretatie van de huidige wetenschappelijke
stand van zaken. Het huidige achtergronddocument
(Alcoholhoudende dranken) vertoont grote discrepanties
met het achtergronddocument Alcohol. Dit is overigens niet
verrassend omdat de epidemiologie niet in staat is
drankspecifieke verschillen goed uit elkaar te trekken.
Redenen hiervoorzijn de verschillen tussen een bier— en
een wijndrinker die veelal niet worden meegewogen (een
belangrijke factor is dieet) en het feit dat mensen
nauwelijks alleen bier of alleen wijn drinken.
We willen met onderstaand commentaar een constructieve
bijdrage leveren aan het document ‘Concept
Achtergronddocument Richtlijnen go ede voeding 2015     —
Alcoholhoudende dranken’. Echter, gezien de genoemde
punten en daarmee de zwakheden in het onderzoek naar
drankspecifieke effecten, adviseren wij dat u in overweging
</pre>

====================================================================== Einde pagina 8 =================================================================

<br><br>====================================================================== Pagina 9 ======================================================================

<pre>neemt om op basis van de huidige wetenschappelijke
kennis geen onderscheid te maken in alcoholhoudende
dranken en het voedingsadvies te richten op
alcoholconsumptie in het algemeen zoals in de vorige
editie van dé Richtlijnen goede voeding (2006) en zoals
ook gedaan wordt in vele andere voedingsadviezen, zoals
bijvoorbeeld die in de Dietary Guidelines for Americans,
2010.
Met vriendelijke groet,
Dr. Ir. Aafje Sierksma
Directeur I(ennisinstituut Bier
Bijlage: Commentaar 1 tlm 3
</pre>

====================================================================== Einde pagina 9 =================================================================

<br><br>====================================================================== Pagina 10 ======================================================================

<pre>Corn mentaar 1 (lnterventieonderzoek):
Ondanks uw bewuste keuze voor de intermedialren (bloeddruk, LDL cholesterol en BMI),
willen wij aangeven dat het in het geval van alcoholconsumptie ook relevant is om te
                              1 c.q. HDL gemedieerde cholesterol efflux
kijken naar HDL cholesterol verhoging,                                         2 en ook
                             36 Daarnaast zijn er nog een aantal andere belangrijke
zijn andere beschermende functies.
factoren die een causaal verband aannemelijk maken, zoals fibrinogeen
                                                                 1 en HbAlc       niet
                                                                                  7
geëvalueerd. In al deze onderzoeken wordt geen onderscheid gevonden tussen
alcoholhoudende dranken, waarmee dus gesuggereerd wordt dat het om een
alcoholeffect gaat.
lnterventieonderzoek maakt het zeer aannemelijk dat er een causaal verband is tussen
matige alcoholconsumptie (dus geen drankspecifieke effecten) en een lagere incidentie
                                                                               1 en
van hart— en vaatziekten, zoals besproken in een systematisch review en meta—analyse
cohort studies.
        6
Wat betreft effect op lichaamsgewicht is er in 2012 een meta—analyse verschenen van
Bendsen en collega’s over de relatie bierconsumptie en obesitas.
                                                            9
Commentaar 2 (Cohortonderzoek):
Bij paragraaf 3.2.1 (Bier) worden in de toelichting niet altijd de juiste percentages
overgenomen. Zo moet in regel 167 ‘ongeveer 5 procent’ vervangen worden door ‘6
procent’ en in regel 174 ‘45 procent’ vervangen door ‘47 procent’ en in regel 180 ‘40 en 85
procent’ vervangen worden door ‘41 en 86 procent’. In het artikel van Ferrari wordt
terecht gewezen op het volgende: “In this study beer use displayed more apparent risk
patterns than wine consumption, particularly in men. Although we believe that this
finding is relevant, we calI for cautious interpretatioYls of these results, as the lifestyle
profile of wine and beer drinkers is profoundly different.” [liervoor verwijzen wij door
naar commentaar3 waarin ingegaan wordt op de eetpatronen van bier—, wijn— en
gedistilleerd drinkers.
Bij paragraaf 3.3.1 (Bier) wordt geconcludeerd dat een verband tussen bierconsumptie
en het risico op hart— en vaatziekten onwaarschijnlijk is, terwijl in paragraaf 3.3.2 (Wijn)
wel uitgebreid in wordt gegaan op de bevindingen uit het onderzoek van Constanzo wat
betreft wijn en hart— en vaatziekten. Constanzo en collega’s schrijven in hun artikel:
“Unfortunately, the very limited data available about either beer or spirit consumption in
relation to carcliovascular or total mortality, did not allow us to perform a fully meta—
analytic investigation on the latter two beverages”. Met dit gegeven is het ons inziens
onredelijk om het verband tussen bierconsumptie en het risico op hart— en vaatziekten als
onwaarschijnlijk aan te duiden. Ook omdat de auteurs in de discussie specifiek aangeven:
“A previous meta—analysis had shown a dear inverse dose—effect curve againstvascular
events for wine but not for beer intake. Evidence from the current updated and extended
 meta—analysis confirms the significant reduction of overall vascular risk associated with
wine consumption and shows, apparently for the first time, a similari—shaped
 relationship between beer intake and cardiovascular risk. Moreover, the comparison of
studies which included a parallel, separate evaluation ofwine and beer consumption,
 indicates a similar protecting effect of either beverage against cardiovascular risk.”
</pre>

====================================================================== Einde pagina 10 =================================================================

<br><br>====================================================================== Pagina 11 ======================================================================

<pre>Bij paragraaf 3.4 (Diabetes Meillitus type 2) wordt ons inziens onterecht geconcludeerd
dat bierdrinkende mannen een hoger risico hebben op diabetes mellitus type 2 dan
mannen die geen bier drinken en dat er geen verband is gevonden bij vrouwen, en dat er
alleen met wijnconsumptie een geringe risicoverlaging is op diabetes mellitus type 2. Wij
worden gesterkt in onze mening door de overall conclusie van dit onderzoek, waarbij
vooral wordt ingegaan op een alcoholeffect en niet drankspecifieke effecten en de
discussie waarin wordt aangegeven dat mogelijk leefstijl (zoals dieet) het verschil verklaart
tussen de bier— en de wijndrinker (zie ook toelichting bij commentaar 3):
Amongst men, moderate alcohol consumption was nonsignificantly associated with a
lower incidence of diabetes with a hazard ratio (H R) of 0.90 (95% Cl: 0.78—1.05) for 6.1—
12.0 versus 0.1— 6.0 g day), adjusted for dietary and diabetes risk factors. Flowever, the
lowest risk was observed at higher intakes of 24.1—96.0 g day) with an HR of 0.86 (95% Cl:
0.75—0.98). Amongstwomen, moderate alcohol consumption was associated with a lower
incidence of diabetes wîth a hazard ratio of 0.82(95% Cl: 0.72—0.92) for 6.1—12.0 g day) (P
interaction gender <0.01).”
The specific risk reduction associated with wine consumption, however, appears to
contradictthefindings of several mechanistic studies. Itwas previously shown thatthe
reduced risk of diabetes with moderate alcohol consumption can be explained by
 increased adiponectin concentrations for 25—30%. However, randomized trials in study
 populations consumîng a variety of alcoholic beverages could not detect a difference in
the effects on adiponectin concentrations. This suggests that the underlying biological
 mechanism is most probably explained by alcohol itself. The specific risk reduction
 observed with wine could thus be attributed to other factors associated with wine
 consumption. Previous studies have shown that wine drinkers differ from drinkers of
 other beverages by consuming a healthier diet and being less likely to smoke. As men and
 women may also differ with regard to such health—related behaviours, as is seen in the
 different structure of confounders amongst men and women, this could in part explain
 the specific association observed for wine consumption and the different effects between
 men and Wo men.”
 Paragraaf 3.6 Borstkanker: Het feit dat er in het achtergronddocument ‘Alcohol’ wel een
 grote bewijskracht wordt gevonden aangaande alcoholconsumptie en risico op
 borstkanker bij vrouwen, terwijl in het achtergronddocument Alcoholhoudende dranken
 voor geen van de alcoholhoudende dranken een eenduidige uitkomst wordt gevonden, is
 tegenstrijdig. Smith—Warner en collega’s geven aan: The specific type of alcoholic beverage
 did not strongly influence risk estimates. Tjønneland en collega’s concluderen: This large
  European study supports previous findings that recent average alcohol intake,
  irrespectively of beverage type, increases the risk of breast cancer.” Deze bevindingen
 sterken nogmaals onze overtuiging dat het gaat om een alcoholeffect en dat daarom een
 voedingsadvies op basis van alcoholconsumptie en niet gespecificeerd per drank de
 voorkeur heeft.
</pre>

====================================================================== Einde pagina 11 =================================================================

<br><br>====================================================================== Pagina 12 ======================================================================

<pre>Commentaar 3: Bier-, wijn- en gedistilleerddrinkers en hun verschillen
Een belangrijke reden om geen onderscheid te maken tussen een bier—, wijn— en
gedistilleerddrinker, is omdat deze nagenoeg niet bestaan. Er wordt nauwelijks alleen bier
of alleen wijn of alleen gedistilleerd gedronken. Dit blijkt ook uit onderzoek van Sluik en
0 Zij hebben deelnemers aan de VCP 2007—2010 ingedeeld naar drankvoorkeur,
collega’s.’
waarbij als criterium is gebruikt dat als 70% van de consumptie bestond uit wijn, dan wel
bier, dan wel gedistilleerd men respectievelijk een wijn, bier—, gedistilleerddrinkeris. Als
het aantal glazen bier, wijn of gedistilleerd niet optelde tot7O%, dan had men geen
voorkeur. Op basis van deze, overigens niet officieel bestaande, definitie waren de
drankvoorkeuren als volgt:
           Blervoorkeur      Wijnvoorkeur        Gedistilleerd voorkeur Geen voorkeur Geen alcohol
Man       32%                10%                 5%                       33%                 20°h
Vrouw 5%                     26%                 6%                       22%                 41%
Verder hebben Sluik en collega’s in hetzelfde onderzoek gekeken naar de eetpatronen van
de op deze manier gedefinieerde bier—, wijn—en gedistilleerd drinkers.’            1 Mensen meteen
voorkeur voor bier hadden ongezondere eetgewoonten dan mensen met een voorkeur
voor wijn. Hierdoor is het dus niet uitgesloten dat dieet een confounder is in de relatie
tussen alcoholconsumptie en gezondheid, waarvoor veelal niet wordt gecorrigeerd. Dat
zou een reden kunnen zijn dat de zogenoemde bierdrinkers er daarom “slechter” vanaf
komen dan de zogenoemde wijndrinkers.
Ook uit een systematische review door Sluik en collega’s blijkt dat drankvoorkeur
gerelateerd is aan eetgewoonten. Zij concluderen dat als er specifiek naar drankvoorkeur
gekeken wordt in relatie tot gezondheid, voeding zeker moet worden meegenomen als
confounder aangezien onderliggende voedingsvoorkeuren vaak eerder gerelateerd zijn aan
gezondheid dan het type drank.
Refe ren ties
1. Brien, S. E., Ronksley, P. E., Turner, B. J., Mukamal, K. J. & Ghall, W. A. Effect of alcohol consumptlon on
blological markers associated wlth risk of coronary heart dlsease: systematic revlew and meta—analysis of
Interventional studies. BMJ (Clinical research ed.) 342, d636, dol:1O.11361bmj.d636 (2011).
2. Rohatgi, A. et al. HDL cholesterol effiux capacity and Incident cardiovascular events. The New England journal
of medicine 371, 2383—2393, doi;1O.1O56INEJMoa14O9O65 (2014).
3. Slerksma, A., van der Gaag, t’l. S., van Tol, A., James, R. W. & Hendriks, H. F. Klnetlcs of HOL cholesterol and
paraoxonase activlty in moderate alcohol consumers. Alcohollsm, clinical and experimental research 26, 1430—
1435, doi:10.1097101.alc.0000030639.57507.60 (2002).
4. Sierksma, A. et al. Effect of moderate alcohol consumptiori on parameters of reverse cholesterol transport in
postmenopausal women. Alcohollsm, cllnical and experimental research 28, 662—666 (2004).
5. van der Gaag, M. S. et al. Dally moderate alcohol consumptlon lncreases serum paraoxonase activity; a diet—
controlled, randomised Intervention study In mlddie—aged men. Atherosclerc,sls 147, 405-410 (1999).
6. van der Gaag, M. S. et al. Alcohol consumptlon stimulates early steps In reverse cholesterol transport. Journal
of lipid research 42, 2077—2083 (2001).
7. Schrieks, l.C. et al. The effect of alcohol consumption on Insulin sensitivity and glycemlc status: a systematlc
review and meta—analysis of intervention studies. Diabetes Care 38, 723—732 (2015).
8. Mukamal, K. J. et al. Drinking frequency, medlatlng biomarkers, and risk of myocardial lnfarction In women
and men. Circulatlon 112, 1406—1413, dol:10.ll6llclrculationaha.105.537704 (2005).
9. Bendsen, N. T. et al. Is beer consumptlon related to measures of abdomlnal and general obesity? A systematic
revlew and meta—analysis. Nutr Rev 71, 67—87, dol:10.1111lj.1753—4887.2012.00548.x (2013).
10. Sluik, 0. et al. Alcoholic beverage preference and diet in a representatlve Dutch population: the Dutch natlonal
food consumption survey 2007—2010. Eur J Clln Nutr 68, 287—294. dol: 10.lO38Iejcn.2013.279 (2014).
11. Slulk, D. et al. Alcoholic Beverage Preference and Dietary Habits: A Systematic Literature        Review. Crit Rev
Food Sci Nutr 15Feb 12:0 (2015).
</pre>

====================================================================== Einde pagina 12 =================================================================

<br><br>====================================================================== Pagina 13 ======================================================================

<pre>M =
i E
un 8 m E
NI = m m
i De gn
u NI ji
E = 7 7 = 8
i
Ty ki
= "Lu 1
E H Ni ;,
a
E = 46 E
=
SOM
ra Bl
[=
on = i
7 ’
7 +
mn nn
4 N E
7 E UE
m Ni E
E E
E
En
[n E E u
ur 1 E
M m u Ni
A
nn E
E E
” * ©
u E
NN n= i E
AA m
2

= a
=o a
MOE VA
7 A
= Te
me E

AN
a a
a
i
a a
Ja

i ER oe
5 1
a: = nn
En me ml
4 7 1
m 7
Ew ©
0 7
5
‘= 7 u
m
M
= 8
N 7
i
TA
7 m nm
7
a. = =
Ää
7 a"
OTT 7
a

4
</pre>

====================================================================== Einde pagina 13 =================================================================

<br><br>====================================================================== Pagina 14 ======================================================================

<pre>  S
  N[D[RLANfl
  INSHRJtJJ
  VOOR AICC)HOLHEWO
Utrecht, 21 augustus 2015
Hierbij ontvangt u de tweede reactie (nu als onderdeel van de vijfde commentaarronde) van het
Nederlands Instituut voor Alcoholbeleid STAP op het concept rapport ‘Achtergronddocument
Richtlijnen goede voeding 2015, deelstudie Alcoholhoudende dranken’.
     1.     Wij hebben ernstige twijfels bij de conclusie in het rapport dat er geen eenduidig verband
            zou bestaan tussen alcoholgebruik en het ontstaan van borstkanker bij vrouwen. In de bijlage
            bij deze reactie sturen we twee recente artikelen mee waarin duidelijke uitspraken worden
            gedaan over de samenhang tussen alcoholgebruik en het ontstaan van borstkanker en
            waarbij geen sprake is van het ontbreken van een eenduidig verband.
     2.     Wat ons verontrust is dat in uw rapport steeds nadrukkelijk onderscheid wordt gemaakt
            tussen wat bekend is over de relatie tussen borstkanker (en ook in geval van andere ziekten)
            en wijngebruik, borstkanker en biergebruik en borstkanker en het gebruik van sterke drank.
            En dat terwijl algemeen bekend is dat het primair gaat om de relatie tussen het gebruik van
            alcohol als carcinogene stof en het ontstaan van diverse ziekten waaronder kanker. Het kan
            niet zo zijn dat we als resultaat van uw rapport gaan zien dat de diverse soorten
            alcoholhoudende drank wat de risico’s van het gebruik ervan betreft, tegen elkaar
            uitgespeeld worden. terwijl de kern is dat de alcohol die deze dranken bevatten als zodanig
            de belangrijkste factor die bepalend is voor de risico’s, Ik vraag u tenminste een apart
            hoofdstuk te wijden aan de samenhang tussen de besproken ziekten en het gebruik van
            alcohol zonder daarbij onderscheid te maken tussen de diverse verschijningsvormen van
            alcohol. Dit mede gezien het feit dat er hoe langer hoe meer tussenproducten
            geconsumeerd worden, zoals Desperados (bier met Tequila), en Muscat (wijn met een
            scheutje gedistilleerde alcohol).
     3.     Ik stuur u nogmaals het CUP rapport van het WCRF over borstkanker en de twee zeer recente
            artikelen van studies die het verband tussen alcohol en borstkanker eenduidig aantonen.
Het betreft de artikelen:
Alcohol intake and breast cancer in the European prospective investigation into cancer and
nutrition
Auteurs: Isabelle Romieu, Chiara Scoccianti, Veronique Chaje’s, Jordi de Batile, Carine Biessy, Laure
Dossus, Laura Baglietto, Francoise Clavel-Chapelon, Kim Overvad, Anja Olsen, Anne Tjønneland,
Rudolf Kaaks, Annekatrin Lukanova, Heiner Boeing, Antonia Trichopoulou, Pagona Lagiou, Dimitrios
Trichopoulos, Domenico Palli, Sabina Sieri, Rosario Tumino, Paolo Vineis, Salvatore Panico, H. B(as)
Bueno-de-Mesquita, Carla H. van Gils, Petra H. Peeters, Eiliv Lund, Gun Skeie, Elisabete Weiderpass,
Jose Ram_on Quiros Garc, Maria-Dolores Chirlaque, Eva Ardanaz, Maria-Jose Sanchez, Eric J. DuelI,
Pilar Amiano, Signe Borgquist, Elisabet Wirf, Goran Hallmans,Ingegerd iohansson,Lena Maria Nilsson,
Kay-Tee Khaw, Nick Wareham, Timothy i. Key, Ruth C. Travis, Neil Murphy, Petra A. Wark, Pietro
Ferrari and Elio Riboli
Gepubliceerd in: Int. J. Cancer: 137, (2015) 1921—1930
</pre>

====================================================================== Einde pagina 14 =================================================================

<br><br>====================================================================== Pagina 15 ======================================================================

<pre>en
Light to moderate intake of alcohol, drinking patterns, and risk of cancer: resuits from two
prospective US cohort studies
Auteurs: Yin Cao, Walter C Willett, Eric B Rimm, Meir i Stampfer, Edward L Giovannucci
Gepubliceerd in: British Medical Journal: (2015), 351-h4238.
Hoogachtend,
Ir. Wim van Dalen
Nederlands Instituut voor Alcoholbeleid STAP
Postbus 9769 1 3506 GT 1 Utrecht
</pre>

====================================================================== Einde pagina 15 =================================================================

<br><br>====================================================================== Pagina 16 ======================================================================

<pre>World Sy American
Cancer wy Institute for

Research Fund uy Cancer Research

Continuous Update Project
Keeping the science current

Breast Cancer
2010 Report

Food, Nutrition, Physical Activity,
and the Prevention of Breast Cancer

Continuous
Update Project

GE P/

</pre>

====================================================================== Einde pagina 16 =================================================================

<br><br>====================================================================== Pagina 17 ======================================================================

<pre>WORLD CANCER RESEARCH FUND GLOBAL NETWORK
OUR VISION
The World Cancer Research Fund global network heips people make choices
that reduce their chances of developing cancer.
OUR HERITAGE
We were the first cancer charity:
•  To create awareness of the relationship between diet and cancer risk
•  To focus funding on research into diet and cancer prevention
•    To consolidate and interpret global research to create a practical
     message on cancer prevention
OUR MISSION
Today the World Cancer Research Fund global network continues:
•    Funding research on the relationship of nutrition, physical activity and
     weight management to cancer risk
•    lnterpreting the accurnulated scientific literature in the tield
•    Educating people aboutchoicestheycan maketo reducetheirchances
     of developirig cancer
THE WCRF GLOBAL NETWORK
The World Cancer Research Fund (WCRF) global network comprises WCRF
International, which operates as the umbrella association for the global
network’s four charitable organisations: The American Institute for Cancer
Research (AICR); World Cancer Research Fund (WCRF UK); World Cancer
Research Fund Netherlands (WCRF NL); World Cancer Research Fund Hong
Kong (WCRF HK).
</pre>

====================================================================== Einde pagina 17 =================================================================

<br><br>====================================================================== Pagina 18 ======================================================================

<pre>Please cite the Report as follows:
World Cancer Research Fund / American Institute for Cancer Research. Continuous
Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of Breast
Cancer. 2010
This Report provides an updated version of section 7.10 Breast Cancer from the Second
Expert Report: Food, Nutrition, Physical Activity and the Prevention of Cancer: a Global
Perspective. This section has been updated with the latest information from the 2008
Continuous Update Project Breast Cancer SLR prepared by a team at Imperial College
London, UK (see acknowledgements). For further details on the epidemiological evidence
please see the full 2008 Continuous Update Project Breast Cancer SLR (Second Epert
Report). For further details on mechanisms please see the Second Expert Report.
The First and Second Expert Reports represent the most extensive analysis of the existing
science on the subject to date. To keep the evidence current and updated into the future,
WCRF/AICR is undertaking the Continuous Update Project, in collaboration with Imperial
College London. The Continuous Update Project builds upon the work conducted for the
Second Expert Report and began by merging all the databases from the different cancer
sites into an upgraded data base.
The Continuous Update Project provides the scientific community with a comprehensive
and up to date depiction of scientific developments on the relationship between diet,
physical activity, obesity and cancer. It also provides an impartial analysis and
interpretation of the data as a basis for reviewing and where necessary revising
WCRF/AICR’s cancer prevention recommendations based on the 2007 Expert Report.
In the same way that the Second Expert Report was informed by a process of systematic
literature reviews (SLRs), the Continuous Update Project systematically reviews the
science. WCRF/AICR has convened a panel of experts (the Continuous Update Project
Panel (see acknowledgements) consisting of leading scientists in the field of diet,
physical activity, obesity and cancer who consider the evidence produced by the
systematic literature reviews and meta-analyses, and consider the results and draw
conciusions before making recommendations.
The updates to the SLRs are being conducted by a team of scientists at Imperial College
London in liaison with the SLR centres where possible.
Instead of periodically repeating the extensive task of conducting multiple systematic
literature reviews that cover a long period of time, the continuous review process is based
on a live system of scientific data that is updated on an ongoing basis from which, at any
point in time, the most current review and meta—analysis of scientific data can be
performed.
Periodically WCRF/AICR will produce reports which will outline the scientific
developments in the field of diet, physical activity, obesity and cancer. The reports may
also include updates to the WCRF/AICR recommendations.
The updated recommendations will be used by the WCRF/AICR education and media
relation departments to inform the general public both of the benefits of a healthy
lifestyle and of the developments in science that underpin these recommendations.
</pre>

====================================================================== Einde pagina 18 =================================================================

<br><br>====================================================================== Pagina 19 ======================================================================

<pre>New information in this report
Section 1. Updated with recent mortality and survival data.
Section 2. Updated section on family history
Section 3. No update
Section 4. No update
Section 5. A new section briefly describing the methodology of the Continuous Update
           Project
Section 6. Evidence has been updated based on the 2008 Continuous Update Project
           Breast Cancer SLR and judgements from the Continuous Update Project
           Panel
Section 7. Provides a comparison with the Second Expert Report.
Since publication of this report in 2011, some changes have been
made to the design and formatting, but no changes have been
made to the content of the report or Panel conciusions. Please
note, however, that the Second Expert Report matrix in this report
has been replaced with the Continuous Update Project Matrix (on
page 3).
                                                                                    2
</pre>

====================================================================== Einde pagina 19 =================================================================

<br><br>====================================================================== Pagina 20 ======================================================================

<pre> F000, NLTFRflION, PHYSICAL ACTIVITV AND
 BREAST CANCER (PHEMENOPALISE) 2010
                      OECREASES RISK                    INGREASES RiSK
                     bctation                  AIc&1ic diinks
                     Body  fatiias             Muk attaired heihtt
                                               reater birtti weight
                     PlysicaI activity
                 1   Dier ibro;        eabI ar: fruita; sc a9d
                     iroota: mat; fish; rnilk an dairy prcL: taI
                     ft; f ao: taTir l; :alc ITi; E              ind&x; dir
                     Lta’i; adult ‘iht         r; b::rira. atrs
                     Nor identifd
1     -zxd r       ..r1i. ait t rwdy Ih fl cn:.r II              :I
  .irmcnI4,
  x4:ir In :rrL:n :4 irI irwt n :hç4 a2 t3—L’:rd :ip:Il iaI.
2Fl         :1 iI ;c o:cuIIr4, tçod, 7ncfl rl :ciixi
 FOOD, NUTRITION, PHYSICAL ACTIVflY AND
 BREAST CANCER (POSTMENOPAUSE) 2010
                      DECREASES RISK                   INCREASES RISK
                     Lactatioa                 Aico1aIi ftirks
                                               Bo.dy fat41es
                                               diiIt attaiai€d igtit’
                     r’hioaI ati’ty            Ah4Jomira1 +atnas
                                               AAJIJIt weigtit gaiii
                                               Total   fat
                     DiLa ftio; vjeLa.2s ard frLi[a: ya and aca
                     prDducs; i fish; rr k ar. dairy prduc; ¶olata;
                     vitanin 0; cIcILn: ac’teriLn: Icaerr r.dax: d .aary
                     pa:erra; bir-Ji wa     enary irtae
                     Ncne idortied
1 ‘:J Iqt’t I u z :PcLy Ic oit ihc rIsk cl ‘cl             mtr ‘0
  T--l. -i r: I: nj1Llcç Ia;I ttcLr vt;                    n Lh pc tm
                              rl ncc rpti2
2 lr il:yul l::                          lr.npqi drcrcncn.l
                                                                            3
</pre>

====================================================================== Einde pagina 20 =================================================================

<br><br>====================================================================== Pagina 21 ======================================================================

<pre>Cancer of the breast is the most common cancer in women woridwide. Around 1.1 million
cases were recorded in 2004.
Observed rates of this cancer increase with industrialisation and urbanisation, and also
with facilities for early detection. It remains much more common in high-income countries
but is now increasing rapidly in middle- and low-income countries, including within Africa,
much of Asia, and Latin America. Breast cancer is fatal in under half of all cases and is
the leading cause of death from cancer in women (fifth for men and women combined),
accounting for 16 per cent of all cancer deaths worldwide in 2004.
Breast cancer is hormone related, and the factors that modify risk of this cancer when
diagnosed premenopausally and when diagnosed postmenopausally (much more
common) are not the same.
The Continuous Update Project Panel judges as follows:
The evidence that lactation protects against breast cancer at all ages is
convincing.
Physical activity probably protects against breast cancer postmenopause,
and there is limited evidence suggesting that it protects against this
cancer diagnosed premenopause. The evidence that alcoholic drinks are a
cause of breast cancer at all ages is convincing. The evidence that the
factors that lead to greater adult attained height, or its consequences, are
a cause of postmenopausal breast cancer is convincing, and these are
probably also a cause of breast cancer diagnosed premenopause.
The factors that lead to greater birth weight, or its consequences, are
probably a cause of breast cancer diagnosed premenopause. Adult weight
gain is probably a cause of postmenopausal breast cancer. The evidence
that body fatness is a cause of postmenopausal breast cancer is
convincing, and abdominal body fatness is probably also a cause. 0fl the
other hand, body fatness probably protects against breast cancer
diagnosed premenopause. There is limited evidence suggesting that total
dietary fat is a cause of postmenopausal breast cancer.
Life events that protect against breast cancer include late menarche, early
pregnancy, bearing children, and early menopause, all of which have the
effect of reducing the number of menstrual cycles, and therefore lifetime
exposure to oestrogen. The reverse also applies.
See chapter 8 of the Second Expert Report for evidence and judgements
on factors that modify risk of body fatness and abdominal fatness,
including physical activity and sedentary ways of life, the energy density of
foods and drinks, and breastfeeding.
In final summary, the strongest evidence, corresponding to judgements of
“convincing” and “probable” show that lactation protects against breast
cancer; that alcoholic drinks are a cause of this cancer; that the factors
that lead to a greater adult attained height, or its consequences, are a
cause of postmenopausal and probably also premenopausal breast cancer;
that factors leading to greater birth weight, or its consequences, are
                                                                                          4
</pre>

====================================================================== Einde pagina 21 =================================================================

<br><br>====================================================================== Pagina 22 ======================================================================

<pre>probably a cause of premenopausal               breast cancer; and that abdominal
body fatness and adult weight                  gain are probably a cause of
postmenopausal          breast     cancer.     Body    fatness      is   a     cause     of
postmenopausal           breast     cancer     but    probably       protects      against
premenopausal breast cancer.
Breast tissue comprises mainly fat, glandular tissue (arranged in lobes), ducts, and
connective tissue. Breast tissue develops in response to hormones such as oestrogens,
progesterone, insulin and growth factors. The main periods of development are during
puberty, pregnancy, and lactation. The glandular tissue atrophies after menopause.
Breast cancers are almost all carcinomas of the epithelial cells lining the ducts (the
channels in the breast that carry milk to the nipple).[1] Premenopausal and
postmenopausal breast cancers are considered separately in this Report. Although rare
(less than 1 per cent of cases [2]), breast cancer can occur in men, but it is not inciuded
here.
1. Trends, incidence, and survival
Breast cancer is the most common cancer in women in high-, middle- and low-income
countries.[3] Age-adjusted rates of breast cancer in women are increasing in most
countries, particularly in areas where the incidence had previously been low, such as
Japan, China and south-eastern and eastern Europe.[4, 5]
This is predominately a disease of high-income countries where overall rates are nearly
three times higher than in middle— to low-income countries. Around the world, age
adjusted incidence rates range from 75-100 per 100 000 women in North America,
northern Europe, and Australia, to less than 20 per 100 000 in parts of Africa and Asia.
[6] In the USA, rates are higher among white women than those from other ethnic groups,
although mortality is highest in black women.[7]
Overall risk doubles each decade until the menopause, when the increase slows down or
remains stable. However, breast cancer is more common after the menopause. Studies
of women who migrate from areas of low risk to areas of high risk assume the rate in the
host country within one or two generations. This shows that environmental factors are
important in the progression of the disease.[8]
Breast cancers can often be detected at a relatively early stage. In countries that provide
or advocate screening, most of these cancers are diagnosed when the disease is still at a
localised stage.[9] Survival rates range from 90 to less than 50 per cent, depending on
the characteristics of the tumour, its size and spread, and the availability of
treatment.[10j Average 5-year survival rates are more than 80% in North America,
Sweden, Japan, Finland and Australia compared with less than 60 per cent in Brazil and
Slovakia and less than 40 per cent in Algeria.[11] The low survival rate in middle- and
low-income countries can be explained mainly by a lack of early detection programmes,
resulting in a high proportion of women presenting with late-stage disease, as well as by a
lack of adequate diagnosis and treatment facilities. Breast cancer accounts for nearly 23
per cent of all cancer incidence in women and 16 per cent of all cancer deaths (all sites
except for skin (non-melanoma) and in women only). [3, 6] Breast cancer is the ninth
most common cause of death in high income countries and around 69% of all breast
cancer deaths occur in middle- and low-income countries.[3] Mortality rates have
remained fairly stable between 1960 and 1990 in most of Europe and the Americas; and
                                                                                          5
</pre>

====================================================================== Einde pagina 22 =================================================================

<br><br>====================================================================== Pagina 23 ======================================================================

<pre> have since showed a decline, which has reached 25-30% in northern Europe.[12] See
 box 1.
Box 1 cancer incidence and survival
The cancer incidence rates and figures given in this Report are those reported by cancer registries, now
established in many countries. These registries record cases of cancer that have been diagnosed. However,
many cases of cancer are not identified or recorded: some countries do not have cancer registries; regions
of some countries have few or no records; records in countries suffering war or other disruption are bound
to be incomplete; and some people with cancer do not consult a physician. Altogether, this means that the
actual incidence of cancer is higher than the figures given here. The cancer survival rates given in this
chapter and elsewhere are usually overall global averages. Survival rates are generally higher in high
income countries and other parts of the world where there are established services for screening and early
detection of cancer and well established treatment facilities. Survival also is often a function of the stage at
which a cancer is detected and diagnosed. The symptoms of some internal cancers are often evident only
at a late stage, which accounts for relatively 10w survival rates. In this context, survival’ means that the
person with diagnosed cancer has not died 5 years after diagnosis.
2. Pathogenesis
Breast tissue, as well as hormones and hormone-receptor status, varies at different
stages of life. It is therefore possible that individual risk factors will have different effects
at different life stages (see 6. Evidence and Judgements). Early menarche, late
menopause, not bearing children, and late (over 30) first pregnancy all increase breast
cancer risk.[8, 13] The age when breasts develop, and menopause, are both influenced
by nutrition, with overnutrition leading to early puberty and late menopause;
undernutrition delays puberty and advances menopause (see chapter 6.2 Second Expert
Report).
Hormones play an important role in breast cancer progression because they modulate
the structure and growth of epithelial tumour cells.[10] Different cancers vary in hormone
sensitivity. Many breast cancers also produce hormones, such as growth factors, that act
locally, and these can both stimulate and inhibit the tumour’s growth.[14, 15]
Family history of breast cancer is associated with a 2-3 fold higher risk of the disease.
Some mutations, particularly in BRCA1, BRAC2 and p53 result in a very high risk of breast
cancer. These mutations are rare and account for only 2 to 5 per cent of total cases.[16]
In addition, growth factor receptor genes, as well as some oncogenes, are overexpressed
in many breast cancers.[10] (Also see box 2.2. chapter 2, Second Expert Report).
3. Other established causes
3.1 General
This section Iists factors outside the scope of this Report, identified as established
causes of cancer by the World Health Organization International Agency for Research on
Cancer, and other authoritative bodies. These factors are listed in Chapter 2.4 of the
Second Expert Report: tobacco use; infectious agents; radiation; industrial chemicals;
and some medications. Other diseases may also increase the risk of cancer. In the same
way, life events that modify the risk of cancer               — causative and protective           —  are also
included.
‘Established’ effectively means beyond reasonable doubt’ roughly the equivalent of the
                                                                         —
judgement of ‘convincing’ used in this Report. Occasionally, authorative findings that
perhaps falI short of ‘established’ are also included here.
                                                                                                               6
</pre>

====================================================================== Einde pagina 23 =================================================================

<br><br>====================================================================== Pagina 24 ======================================================================

<pre>Where possible, a note of interactive or multiplicative effects with food, nutrition, and the
other factors covered by this Report is added, as is any indication of scale or relative
importance. The factors here are almost all causative, whereas much of the evidence on
food, nutrition, physical activity, and related factors shows or suggests protection against
cancer.
3.2 Specific
Life events. Lifetime exposure to oestrogen, influenced by early menarche, late natural
menopause, not bearing children, and late (over 30) first pregnancy all increase the risk
of, and may be seen as causes of, breast cancer.[8, 13] The reverse also applies: late
menarche, early menopause, bearing children, and early pregnancy all reduce the risk of,
and may be seen as protective against breast cancer. Age of breast development and
menopause are influenced by nutrition, with high-energy diets promoting earlier puberty
and late menopause, and lowenergy diets delaying puberty and advancing menopause.
Radiation. lonising radiation exposure from medical treatment such as X-rays,
particularly during puberty, increases risk, even at low doses.[17]
Medication. Hormone replacement therapy is a cause of breast cancer. The increased
risk appears to disappear a few years after cessation.[18] Oral contraceptives containing
both oestrogen and progesterone cause a small, transient, increased risk of breast
cancer; the increased risk disappears after cessation.[19]
4. Interpretation of the evidence specific to breast cancer
4.1 General
For general considerations that may affect interpretation of the evidence, see chapters
3.3 and 3.5, and boxes 3.1, 3.2, 3.6 and 3.7 of the Second Expert Report.
‘Relative risk’ is used in this Report to denote ratio measures of effect, including ‘risk
ratios’, ‘rate ratios’, ‘hazard ratios’, and ‘odds ratios’.
4.2 Specific
Considerations specific to breast cancer include:
Patterns. The preponderance of data from high-income countries is a special issue with
breast cancer. Breast cancer is hormone related, and factors that modify risk have
different effects on cancers diagnosed pre- and postmenopause.
Classification. Because of the importance of menopause as an effect modifier, studies
should stratify for menopause status. Many do not.
Confounding. Hormone replacement therapy is an important possible confounder in
postmenopausal breast cancer. A few studies also reported resuits separately for
different hormone receptor profiles within cancers. High-quality studies adjust for age,
number of reproductive cycles, age at which children were bom, and the taking of
hormone-based medications.
                                                                                            7
</pre>

====================================================================== Einde pagina 24 =================================================================

<br><br>====================================================================== Pagina 25 ======================================================================

<pre>Effect modification. There is growing evidence that the impact of dietary exposures on
risk of breast cancer may differ according to the particular molecular subtypes of cancer.
5. Methodology
To ensure consistency with evidence collected and analysed for the Second Expert Report
much of the methodology following for the Continuous Update Project remains
unchanged from that used previously. Based upon the experience of conducting the
systematic literature reviews for the Second Expert Report some modifications to the
methodology were made. The literature search was restricted to Medline and iricluded
only randomised controlled trials, cohort and case-control studies. The 2008 Continuous
Update Project Breast Cancer SLR inciuded studies published up to December 2007.
Publications in foreign languages were not included. Due to the large number of cohort
studies, analysis and interpretation of oase-control studies was not included in the
Continuous Update Project SLR. Meta-analyses and forest plots of highest versus lowest
categories were prepared for breast cancer incidence. Studies with mortality endpoints
previously included in analyses were removed. Studies reporting mean difference as a
measure of association are not included in the Continuous Update Project SLR as relative
risks estimated from the mean differences are not adjusted for possible confounders,
and thus not comparable to adjusted relative risks from other studies. (For more
information on methodology see the full 2008 Continuous Update Project Breast Cancer
SLR (Second Expert Report).
6. Evidence and judgements
The updated search identified 81 new articles, giving a total of 954 publications for
breast cancer. The following sections include evidence from case-control studies
considered as part of the Second Expert Report; however as mentioned in the previous
section the evidence from oase-control studies was not included in the 2008 Continuous
Update Project Breast Cancer SLR. Fuller summaries of the experimental and
mechanistic evidence can be found in chapters 4-6 of the Second Expert Report. For
information on the criteria for grading the evidence see box 3.8 of the Second Expert
Report. References to studies added in the Continuous Update Project have been
included in the following sections; for details on references to other studies see Second
 EçprtRçpp rt.
6.1 Alcoholic drinks
(Also see sections 3.7.1 Alcoholic drinks and 5.4 Alcohol (as ethanol) of the 2008
Continuous Update Project Breast Cancer SLR)
The Continuous Update Project identified 4 new cohort studies[20-23] that investigated
alcoholic drinks and 2 new cohort studies[24, 25] and 3 recent publications from
previously included cohort studies[26-28] that investigated ethanol intake. For
premenopausal breast cancer a total of 4 cohort studies investigated alcoholic drinks
and 6 cohort studies investigated ethanol intake. The respective numbers for
postmenopausal breast cancer were 9 and 16. For all-age breast cancer a total of 13
cohort studies investigated alcoholic drinks and 11 cohort studies investigated ethanol
intake. Most studies showed increased risk with increased intake. Meta-analysis of cohort
studies for the Second Expert Report showed a 10 per cent increased risk for all-age
breast cancer, a 9 per cent increased risk for premenopausal breast cancer and a 8 per
cent increased risk for postmenopausal breast cancer per 10 g ethanol (Page 167
Second Expert Report). An updated meta-analysis for postmenopausal breast cancer
                                                                                           8
</pre>

====================================================================== Einde pagina 25 =================================================================

<br><br>====================================================================== Pagina 26 ======================================================================

<pre>showed an 8 per cent increased risk per 10 g ethanol (Figure Al 2008 Continuous
Update Project Breast Cancer SLR). The Second Expert Report inciuded 31 case-control
studies that investigated alcoholic drinks and 29 case-control studies that investigated
ethanol intake and all-age breast cancer. Meta-analysis of case-control data showed a 5
per cent increased risk per 5 drinks/week, and a 6 per cent increased risk per 10 g
ethanol/day (Pages 166-167 Second Expert Reporti. Menopausal status did not
significantly alter the association.
Two pooled analyses also showed statistically significant increased risks of 9 and 7 per
cent per 10 g ethanol/day. The first was based on 6 cohort studies with more than 320
000 participants, followed up for up to 11 years, with more than 4300 breast cancer
cases. The other analysed 53 case-control studies, with more than 58 000 cases and
more than 95 000 controls.[29, 30] A meta-analysis of 3 cohort and 7 case-control
studies assessed the association between alcohol intake and the risk of ER-/PR-defined
breast cancer. [31] The dose-response meta-analysis showed that an increase in alcohol
consumption of 10 g of ethanol per day was associated with statistically significant
increased risks for all ER+ (12 per cent), all ER- (7 per cent), ER+PR+ (11 per cent) and
ER+PR- (15 per cent), but not ER-PR-. A statistically significant heterogeneity of the
results across all ER+ versus ER-PR- was observed.
                                                                                        9
</pre>

====================================================================== Einde pagina 26 =================================================================

<br><br>====================================================================== Pagina 27 ======================================================================

<pre>Reactive metabolites of alcohol, such as acetaldehyde, may be carcinogenic.
Additionally, the effects of alcohol may be mediated through the production of
prostaglandins, lipid peroxidation, and the generation of free radical oxygen
species. Alcohol also acts as a solvent, enhancing penetration of carcinogens into
cells. High consumers of alcohol may have diets deficient in essential nutrients,
making tissues susceptible to carcinogenesis. In addition, most experimental
studies in animals have shown that alcohol intake is associated with increased
breast cancer risk. Alcohol interferes with oestrogen metabolism and action in
multiple ways, influencing hormone levels and oestrogen receptors.
There is an interaction between folate and alcohol affecting breast cancer risk:
increased folate status partially mitigates the risk from increased alcohol
consumption.[32]
 The evidence added for the Continuous Update Project is consistent with that
 from the Second Expert Report There is ample and generally consistent
 evidence from cohort and case-control studies.
 A dose-response relationship is apparent. There is robust evidence for
 mechanisms operating in humans. The conclusion reached for the Second
 Expert Report remains unchanged. The evidence that alcoholic drinks are a
 cause of premenopausal and postmenopausal breast cancer is convincing. No
 threshold was identified.
6.2 Lactation
(Also see section 1.6.1 Breastfeeding of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 2 new cohort studies[33, 34] that
investigated ever having breastfed as compared with never having breastfed and
3 new cohort studies[20, 21, 33] that investigated the total duration of lactation.
For each of premenopausal and postmenopausal breast cancer a total of 2 cohort
studies investigated ever having breastfed compared to never having breastfed
and 2 cohort studies investigated total duration of lactation. For all-age breast
cancer 3 studies investigated ever having breastfed and 6 studies investigated
total duration of lactation. The Second Expert Report included 37 case-control
studies that investigated ever having breastfed as compared with never having
breastfed and 55 case-control studies that investigated the total duration of
lactation. Most cohort and case-control studies reported decreased risk with ever
having breastfed and with increasing duration of breastfeeding. Previous meta
analyses from the Second Expert Report for case-control data showed a 2 per
cent decreased risk per 5 months of total breastfeeding; and for cohort data
showed a non-significant decreased risk (Page 241 Second Expert Report).
Pooled analysis from 47 epidemiological studies in 30 countries (more than 50
000 controls and nearly 97 000 breast cancer cases) showed a statistically
significant decreased risk of breast cancer of 4.3 per cent for each 12 months of
breastfeeding. Menopause status was not an effect modifier.[35] The relationship
between breastfeeding and breast cancer according to hormone receptor status
was investigated in a meta-analysis of 5 population-based case-control studies. A
statistically significantly lower risk was found, both of ER+/PR+ breast cancers
</pre>

====================================================================== Einde pagina 27 =================================================================

<br><br>====================================================================== Pagina 28 ======================================================================

<pre>(22 per cent) and for ER-/PR- cancers (26 per cent), for more than 6 months of
breastfeeding compared with never breastfeeding. [36]
Lactation is associated with increased differentiation of breast ceils and with
lower exposure to endogenous sex hormones during amenorrhea accompanying
lactation. in addition, the strong exfoliation of breast tissue during lactation, and
the massive epithelial apoptosis at the end of lactation, could decrease risk by
elimination of ceils with potential DNA damage.
 The evidence added for the Continuous Update Project is consistent with that
 from the Second Expert Report. There is abundant epidemiological evidence
 from both cohort and case-control studies, which is consistent and shows a
 dose-response relationship. There is robust evidence for plausible mechanisms
 that operate in humans. The conclusion reached for the Second Expert Report
  remains unchanged. The evidence that lactation protects against both
  premenopausal and postmenopausal breast cancer is convincing.
6.3 Physical activity
(Also see section 6. Physical Activity of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 2 new cohort studies[37, 38]
investigating total physical activity; 1 new cohort study investigating occupational
activity[37]; 3 new cohort studies[37-39] and 1 recent publication from a
previously included cohort study[40] investigating recreational activity; and 2 new
cohort studies[37, 38] investigating household activity. For premenopausal breast
cancer a total of 5 cohort studies investigated total physical activity and 4, 3 and
1 studies investigated occupational, recreational and household activities
respectively. For postmenopausal breast cancer 2 studies investigated total
activity and 5, 11 and 1 studies investigated occupational, recreational and
household activities respectively. For all-age breast cancer 4 studies investigated
total physical activity and 4, 5 and 1 studies investigated occupational,
recreational and household activities respectively. The Second Expert Report
inciuded 8 case-control studies that investigated total physical activity, 7 case
control studies that investigated occupational activity and 11 case-control studies
that investigated recreational activity.
Menopause age unspecified
Most studies showed decreased risk with increased physical activity. Meta
analysis of case-control studies for the Second Expert Report showed a 10 per
cent decreased risk per 7 MET-hours recreational activity/ week (Page 204
Second Expert Report).
P rem en opa u s e
Data were inconsistent for cohort studies for physical activity; however most case
control studies reviewed for the Second Expert Report showed evidence of
decreased risk (Page 204 Second Expert Report).
Postmenopause
Nearly all of the cohort studies showed decreased risk with increased physical
activity. The meta-analyses from the Second Expert Report of cohort and case
                                                                                    II
</pre>

====================================================================== Einde pagina 28 =================================================================

<br><br>====================================================================== Pagina 29 ======================================================================

<pre>control data both showed a 3 per cent decreased risk per 7 MET-hours
recreational activity/week (Page 205 Second Expert Report).
Sustained moderate physical activity raises the metabolic rate and increases
maximal oxygen uptake. In the long term, regular periods of such activity increase
the body’s metabolic efficiency and capacity (the amount of work that it can
perform), as well as reducing blood pressure and insulin resistance. In addition, it
decreases levels of oestrogens and androgens in postmenopausal women. Some
trials have also shown decreases in circulating oestrogens, increased menstrual
cycle Iength, and decreased ovulation in premenopausal women with a high level
of physical activity.
  Premenopause: There is ample evidence from prospective studies, but it is
  inconsistent. There is evidence from case-control studies suggestive of a
  decreased risk with higher levels of physical activity. The conclusion reached for
  the Second Expert Report remains unchanged. There is limited evidence
  suggesting that physical activity protects against premenopausal breast cancer.
  Postmenopause: The evidence added in the Continuous Update Project is
  consistent with that from the Second Expert Report. There is ample evidence
  from prospective studies showing lower risk of postmenopausal breast cancer
  with higher levels of physical activity, with a dose-response relationship, although
  there is some heterogeneity. There is little evidence on frequency, duration, or
  intensity of activity. The conclusion reached for the Second Expert Report
  remains unchanged. There is robust evidence for mechanisms operating in
  humans. Physical activity probably protects against postmenopausal breast
  cancer.
6.4 Body fatness
(Also see section 8.1.1 Body Mass Index of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 10 new[34, 41-491 and 2 recent
publications from previously inciuded studies[39, 501 investigating body fatness,
as measured by BMI for pre- and postmenopausal breast cancer. For
premenopausal breast cancer there was a total of 22 studies and for
postmenopausal breast cancer there were 28 studies. The Second Expert Report
inciuded more than 100 case-control studies that investigated body fatness.
When grouped for all ages the Second Expert Report showed that the data were
inconsistent in relationship to body fatness (Page 218 Second Expert Report) and
this remained true for the Continuous Update Project. However, a consistent
effect emerged when they were stratified according to menopausal status.
P rem en opa u s e
Most studies showed a decreased risk for premenopausal breast cancer. Meta
analyses for the Second Expert Report (Page 221 Second Expert Report) showed
a 15 per cent decreased risk per 5kg/m  2 for cohort studies and an 8 per cent
                      2 for case-control studies; the updated meta-analysis for
decreased risk per 5kg,/m
                                                                                     12
</pre>

====================================================================== Einde pagina 29 =================================================================

<br><br>====================================================================== Pagina 30 ======================================================================

<pre> cohort studies showed a 7 per cent decreased risk per 5kg/m
                                                         2 (Figure BMI4 2008
 Continuous Update Project Breast Cancer SLR). A pooled analysis of four cohort
studies with 723 cases of premenopausal breast cancer followed up for up to 11
years showed a 14 per cent decreased risk per 5kg/m .[51] A meta-analysis of
                                                    2
 20 cohort studies reported an 8 per cent decreased risk per 5kgJm
                                                             .[52]
                                                             2
 Postmenopause
 Most studies showed an increased risk for postmenopausal breast cancer with
 increased body fatness. Meta-analysis of cohort studies for the Second Expert
 Report (Page 219 Second Expert Report) showed an 8 per cent increased risk per
2 and a 13 per cent increased risk per 5kg/m
5kg/m                                           ; the updated meta-analysis
                                                2
of cohort studies showed a 13 per cent increased risk per 5kg/m
                                                             2 (Figure BMI7
2008 Continuous Update Project Breast Cancer SLR). A pooled analysis of seven
cohort studies with 3208 cases of postmenopausal breast cancer followed up for
up to 11 years showed a 9 per cent increased risk per 5kg/m  .[51] A meta
                                                             2
analysis of 31 cohort studies reported a 12 per cent increased risk per
      . [52]
5 kg/ m
      2
Body fatness directly affects levels of many circulating hormones, such as insulin,
insulin-like growth factors, and oestrogens, creating an environment that
encourages carcinogenesis and discourages apoptosis (programmed ceIl death).
It also stimulates the body’s infiammatory response, which may contribute to the
initiation and progression of several cancers (see chapter 2.4.1.3 Second Expert
Report). Adjusting for serum levels of oestradiol diminishes or destroys the
association with BMI, suggesting that hormones are a predominant
mechanism.[53]
There is no single well established mechanism though which body fatness could
prevent premenopausal breast cancer. According to the oestrogen plus
progesterone theory, overweight premenopausal women would be protected
because they would be more frequently anovulatory, and therefore less likely to
be exposed to endogenous progesterone. However, this theory is not well
supported by recent studies, which suggest that natural progesterone could be
protective.[54] Normal levels of natural progesterone are likely to be protective,
and well nourished, or perhaps overnourished women, who may become slightly
overweight in adulthood, may be protected by their natural fertile condition.
Another possible mechanism is that the increased adipose tissue-derived
oestrogen levels in overweight children could induce early breast differentiation
and eliminate some targets for malignant transformation.[55] Anovulation and
abnormal hormone profiles are commonly associated with obesity.[56] The age
specific pattern of association of breast cancer with BMI, therefore, is largely
explained by its relationship with endogenous sex hormone levels.
Breast cancer diagnosed after the menopause is much more common. Therefore,
throughout life, a decreased risk of premenopausal breast cancer would be
outweighed by an increased risk of postmenopausal breast cancer.
                                                                                 13
</pre>

====================================================================== Einde pagina 30 =================================================================

<br><br>====================================================================== Pagina 31 ======================================================================

<pre> Premenopause: The evidence added in the Continuous Update Project is
 consistent with that from the Second Expert Report. There is a substantial
 amount of consistent evidence epidemiological evidence with a dose-response
 relationship, but the mechanistic evidence is speculative. The conciusion
 reached for the Second Expert Report remains unchanged. Greater body fatness
 probably protects against premenopausal breast cancer.
 Postmenopause: The evidence added in the Continuous Update Project is
 consistent with that from the Second Expert Report. There is abundant and
 consistent epidemiological evidence and a dear doseresponse relationship with
 robust evidence for mechanisms operating in humans. The conciusion reached
 for the Second Expert Report remains unchanged. The evidence that greater
 body fatness is a cause of postmenopausal breast cancer is convincing.
6.5 Adult attained height
(Also see section 8.3.1 Height of the 2008 Continuous Update Project Breast
Cancer SLR)
The Continuous Update Project identified 5 new cohort studies[34, 39, 41, 48,
57] that investigated adult attained height. The total number of cohort studies
was 21 for all-age or age unspecified, 17 for premenopausal and 22 for
postmenopausal breast cancer. The Second Expert Report included 29 case
control studies that investigated adult attained height and all-age breast cancer,
38 for premenopausal and 34 for postmenopausal breast cancer.
Menopausal age unspecified
Most of the studies showed increased risk. Meta-analysis for the Second Expert
Report showed a 9 per cent increased risk per 5cm of height for cohort studies
and a 3 per cent increased risk per 5cm of height for case-control studies (Page
233 Second Expert Report).
Pre me no pa use
Most of the studies showed increased risk. Meta-analysis for the Second Expert
Report showed a 9 per cent increased risk per 5cm of height for cohort studies
and a 4 per cent increased risk per 5cm for case-control studies (Page 235
Second Expert Report). An updated meta-analysis of cohort studies also showed a
9 per increased risk per 5cm of height (Figure Htl 2008 Continuous Update
Project Breast Cancer SLR). A pooled analysis of four cohort studies with 723
cases of premenopausal breast cancer followed up for up to 11 years showed a
non-significant increased risk with greater adult attained height.[51]
Postmenopause
Nearly all the cohort studies and most case-control studies showed increased
risk, with no studies showing statistically significant contrary results. Meta
analyses for the Second Expert Report showed an 11 per cent increased risk per
5cm of height for cohort studies and a 2 per cent increased risk per 5cm for case
control studies (Page 234 Second Expert Report). An updated meta-analysis
showed a 10 per increased risk per 5cm of height (Figure Ht4 2008 Continuous
Update Project Breast Cancer SLR. A pooled analysis of seven cohort studies with
                                                                                14
</pre>

====================================================================== Einde pagina 31 =================================================================

<br><br>====================================================================== Pagina 32 ======================================================================

<pre>3208 cases of postmenopausal breast cancer followed up for up to 11 years
showed a significantly significant 7 per cent increased risk per 5cm of height.[51]
The general mechanisms through which the factors that lead to greater adult
attained height, or its consequences, could plausibly influence cancer risk are
outlined in chapter 6.2.1.3 and box 2.4 of the Second Expert Report. Many of
these, such as early-life nutrition, altered hormone profiles, and the rate of sexual
maturation, could plausibly increase cancer risk.
  Premenopause: There are fewer data for premenopausal than for
  postmenopausal breast cancer. The evidence added in the Continuous Update
  Project is consistent with that from the Second Expert Report. The
 epidemiological evidence is generally consistent, with a dose-response
  relationship and evidence for plausible mechanisms. The conciusion reached for
 the Second Expert Report remains unchanged. The factors that lead to greater
 adult height, or its consequences, are probably a cause of premenopausal
  breast cancer. The causal factor is unlikely to be tallness itself, but factors that
  promote linear growth in childhood.
  Postmenopause: The evidence added in the Continuous Update Project is
 consistent with that from the Second Expert Report. There is abundant
 epidemiological evidence, which is generally consistent, with a dear dose
  response relationship and evidence for plausible mechanisms operating in
  humans. The conclusion reached for the Second Expert Report remains
  unchanged. The evidence that the factors that lead to greater adult attained
  height, or its consequences, are a cause of postmenopausal breast cancer is
 convincing. The causal factor is unlikely to be tallness itself, but factors that
  promote linear growth in childhood.
6.6 Abdominal fatness (postmenopause)
(Also see sections 8.2.1 Waist Circumference and 8.2.3. and Waist to hip ratio of
the 2008 Continuous Update Project Breast Cancer SLR)
The Continuous Update Project identified 3 new cohort studies[42, 47, 48] and 1
recent publication from a previously inciuded cohort study[58] that investigated
waist circumference and 3 cohort studies[42, 47, 48] and 2 recent publications
from previously inciuded cohort studies[28, 59] that investigated waist to hip
ratio. In total 9 cohort studies investigated waist circumference and 13 waist to
hip ratio. The Second Expert Report inciuded 3 case-control studies that
investigated waist circumference and 8 that investigated waist to hip ratio.
All of the waist circumference studies and most of those on waist to hip ratio
showed increased risk with increased measures of abdominal fatness. Meta
analysis of cohort studies for the Second Expert Report showed a 5 per cent
increased risk per 8 cm in waist circumference (Page 226 Second Expert Report).
The updated meta-analyses were stratified by whether the study adjusted for BMI.
Studies that did not adjust for BMI showed a 7 per cent increased risk per 8cm in
waist circumference and those that did showed a 4 per cent increased risk
(Figures W5 and W6 2008 Continuous Update Project Breast Cancer SLR).
                                                                                    15
</pre>

====================================================================== Einde pagina 32 =================================================================

<br><br>====================================================================== Pagina 33 ======================================================================

<pre>Meta-analysis of cohort studies for the Second Expert Report showed a 19 per
cent increased risk per 0.1 increment in waist to hip ratio (Page 226 Second
Expert Report). The updated meta-analyses were stratified by whether the study
adjusted for BMI. Studies that did not adjust for BMI showed a 9 per cent
increased risk per 0.1 increment in waist to hip ratio and those that did showed a
non-significant increased risk (Figures WHR6 and WHR7 2008 Continuous Update
Project Breast Cancer SLR).
The general mechanisms through which abdominal fatness could plausibly cause
cancer are outlined in chapter 6.1.3 9 and box 2.4 of the Second Expert Report.
The hormonal and other biological effects of being overweight or obese are
outlined in chapter 8 of the Second Expert Report. Many of these, such as
increased levels of circulating oestrogens and decreased insulin sensitivity, are
associated with abdominal fatness independently of overall body fatness.
 The evidence added in the Continuous Update Project is consistent with that
 from the Second Expert Report. There is a substantial amount of epidemiological
 evidence but some inconsistency. There is robust evidence for mechanisms that
 operate in humans. The conclusion reached for the Second Expert Report
  remains unchanged. Abdominal fatness is a probable cause of postmenopausal
  breast cancer.
6.7 Adult weight gain (postmenopause)
(Also see section 8.1.6 Weight Change of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 3 new cohort studies[42, 47, 48] and 1
recent publication from a previously included cohort study[60] that investigated
adult weight change and postmenopausal breast cancer. In total 10 cohort
studies investigated adult weight change. The Second Expert Report included 17
case-control studies that investigated adult weight change. Nearly all the studies
showed increased risk with increased weight gain in adulthood. Meta-analyses for
the Second Expert Report showed a 3 per cent increased risk per 5kg gained for
the cohort studies and a 5 per cent increased risk per 5kg for case-control studies
(Page 227 Second Expert Report). Heterogeneity may be explained by failure to
separate postmenopausal women taking hormone replacement therapy.
 Body fatness directly affects levels of many circulating hormones, such as insulin,
insulin-like growth factors, and oestrogens, creating an environment that
encourages carcinogenesis and discourages apoptosis (see chapter 2.7.1.3
Second Expert Report). It also stimulates the body’s inflammatory response,
which may contribute to the initiation and progression of several cancers.
  The evidence added in the Continuous Update Project is consistent with that
  from the Second Expert Report. There is ample, consistent epidemiological
  evidence and a dose-response relationship was apparent. The conciusion
  reached for the Second Expert Report remains unchanged. Adult weight gain is a
  probable cause of postmenopausal breast cancer.
                                                                                  16
</pre>

====================================================================== Einde pagina 33 =================================================================

<br><br>====================================================================== Pagina 34 ======================================================================

<pre>6.8 Greater birth weght (premenopause)
(Also see section 8.4.1 Birthweight of the 2008 Continuous Update Project Breast
Cancer SLR)
The Continuous Update Project identified 1 new cohort study{61] that
investigated birth weight and premenopausal breast cancer. In total 6 cohort and
4 case-control studies investigated birth weight. All cohort studies and most case
control studies showed increased risk with greater birth weight. Meta-analysis of
cohort studies for the Second Expert Report showed an 8 per cent increased risk
per kg (Page 238 Second Expert Report).
The general mechanisms through which the factors that lead to greater birth
weight, or its consequences, could plausibly influence cancer risk are outline in
chapter 6.2.11. of the Second Expert Report many of these, such as long-term
programming of hormonal systems, could plausibly increase cancer risk. Greater
birth weight raises circulating maternal oestrogen levels and may increase insulin
like growth factor (IGF)-1 activity; low birth weight raises fetal and maternal levels
of IGF-1 binding protein. The action of both oestrogens and IGF-1 are thought to
be important in fetal growth and mammary gland development, and play a
central, synergistic role in the initiation and promotion of breast cancer.[62]
Animal experiments also provide evidence that exposure to oestrogens during
fetal and early postnatal development can increase the risk of mammary
cancers.[63]
  The evidence added in the Continuous Update Project is consistent with that
  from the Second Expert Report. There is general consistency amongst the
  relatively few epidemiological studies, with some evidence for a dose-response
  relationship. The mechanistic evidence is speculative. The conclusion reached
  for the Second Expert Report remains unchanged. The factors that lead to
  greater birth weight, or its consequences, are probably a cause of
  premenopausal breast cancer.
6.9 Total fat (postmenopause)
(Also see section 5.2 Total Fat of the 2008 Continuous Update Project Breast
Cancer SLR)
The Continuous Update Project identified 1 new cohort study{64] and 1 recent
publication from a previously inciuded cohort study[65j that investigated total fat
intake and 1 new cohort study{66] and 1 recent publication from a previously
included cohort study[67] that investigated energy from fat and postmenopausal
breast cancer. In total 9 cohort studies investigated total fat intake and 5 cohort
studies investigated energy from fat and postmenopausal breast cancer. The
Second Expert Report included 16 case-control studies that investigated total fat
intake and postmenopausal breast cancer. For total fat most studies showed
increased risk with increased intake. Meta-analyses for the Second Expert Report
showed a non-significant increased risk for cohort studies and an 11 per cent
increased risk per 20g/day for case-control studies (Page 138 Second Expert
Report). A pooled analysis of cohort studies of more than 7300 cases of breast
cancer showed an overall non-significant decreased risk with increased fat intake.
Menopausal status did not significantly alter the result.[68] For energy from fat
                                                                                    17
</pre>

====================================================================== Einde pagina 34 =================================================================

<br><br>====================================================================== Pagina 35 ======================================================================

<pre>most cohort studies reported decreased risk with increasing per cent energy from
fat and one reported a statistically significant increased risk.
The Women’s Health Initiative Dietary Modification Randomised Controlled Trial
with 655 cases of postmenopausal breast cancer reported a relative risk of 0.91
(0.83-1.01) for intervention and comparison group after 8.1 years.[69] Adjusting
for change in body weight had no effect on the relative risk. The trial was
designed to reduce fat intake to 20% and increase servings of vegetables and
fruit to 5 per day and increase servings of grains to at least 6 per day. However for
women with at least 36.8% energy from fat at baseline a decrease was observed
for intervention compared with control (RR- 0.78 (0.64-0.95)).
Higher endogenous oestrogen levels after menopause are a known cause of
breast cancer.[53, 70] Dietary fat may also increase endogenous oestrogen
production.[71]
 The evidence added in the Continuous Update Project is consistent with that
 from the Second Expert Report. Evidence from prospective epidemiological
 studies of different types on the whole shows inconsistent effects, whfle case
 control studies show a significant positive association. Mechanistic evidence is
 speculative. The conciusion reached for the Second Expert Report remains
  unchanged. Overall, there is limited evidence suggesting that consumption of
 total fat is a cause of postmenopausal breast cancer.
6.10 Other exposures
For pre- and postmenopausal breast cancer, other exposures were evaluated.
However, the data were either of too low quality, too inconsistent, or the number
of studies too few to allow conclusions to be reached. The list of exposures is
shown in the matrices under limited no conclusion. Additional meta-analyses of
                                        -
cohort studies on dietary fibre and highest versus lowest category forest plots for
total, red and processed meat, fish, dietary folate and energy were also
conducted as part of the Continuous Update Project (See 2008 Continuous
Update Project Breast Cancer SLR for details).
There is considerable speculation around a biologically plausible interaction of
soy and soya products with breast cancer development, due to their high
phytoestrogen content. Data on pulses (legumes) were sparse and inconsistent.
A meta-analysis of 3 cohort and 6 case-control studies showed a statistically
significant 25 per cent lower risk of breast cancer at any age for highest versus
lowest intake of soy products. [72]
A meta-analysis of 6 cohort and 12 case-control studies reported a statistically
significant 14 per cent lower risk of breast cancer at any age for highest versus
lowest consumption of soy protein (estimated from intake of soy food and dietary
isoflavones). [73] Another meta-analysis reported a statistically significant 12 per
cent lower risk of breast cancer at any age for highest versus lowest intake of
isoflavones.[74] In a subgroup analysis the association was statistically significant
for Asian populations (29 per cent lower risk) but not for Western populations.
[74] These meta-analyses are limited by the difficulty in the standardisation of
                                                                                   18
</pre>

====================================================================== Einde pagina 35 =================================================================

<br><br>====================================================================== Pagina 36 ======================================================================

<pre>measure of soy intake. The quantity and type of soy consumed varied greatly
across the studies, such that the contrasts in intake levels for the reported risk
estimates differed widely. Although resuits of these meta-analyses suggest that
soy intake is associated with a modest reduction in breast cancer risk,
heterogeneity across studies limits the ability to interpretthe findings.
7. Comparison with the Second Expert Report
Overall the evidence from the additional cohort studies identified in the
Continuous Update Project was consistent with those reviewed as part of the
Second Expert Report. Much of the new evidence related to body fatness,
abdominal fatness and weight gain; there were also new studies reporting on
alcohol consumption.
8. Conciusions
Since the new evidence that was found as part of the Continuous Update Project
is consistent with the evidence presented in the Second Expert Report the
conclusions are unchanged.
The Continuous Update Project Panel concludes:
The evidence that lactation protects against breast cancer at all ages thereafter is
convincing. Physical activity probably protects against postmenopausal breast
cancer, and there is limited evidence suggesting that it protects against
premenopausal breast cancer. The evidence that alcoholic drinks are a cause of
breast cancer at all ages is convincing. The evidence that the factors that lead to
greater attained adult height or its consequences are a cause of postmenopausal
breast cancer is convincing; these are probably a cause of premenopausal breast
cancer.
The factors that lead to greater birth weight or its consequences are probably a
cause of breast cancer diagnosed premenopause. Adult weight gain is probably a
cause of postmenopausal breast cancer. The evidence that body fatness is a
cause of postmenopausal breast cancer is convincing, and abdominal body
fatness is probably a cause of this cancer. On the other hand, body fatness
probably protects against breast cancer diagnosed premenopause. There is
limited evidence suggesting that total dietary fat is a cause of postmenopausal
breast cancer.
                                                                                  19
</pre>

====================================================================== Einde pagina 36 =================================================================

<br><br>====================================================================== Pagina 37 ======================================================================

<pre>Acknowledgements
Continuous Update Project Panel Members
      Elisa V Bandera MD PhD
      The Cancer Institute of New Jersey
      New Brunswick, NJ, USA
      David Hunter MBBS ScD
      Harvard University of Public Health
      Boston, MA, USA
      Alan Jackson CBE MD FRCP
      University of Southampton, UK
      John Milner PhD
      National Cancer Institute
      Rockville MD, USA
      Hilary J Powers PhD RNutr
      University of Sheffield, UK
      ArthurSchatzkin MD DrPH
      National Cancer Institute
      Rockville, MD, USA
      Ricardo Uauy MD PhD
      Instituto de Nutricion y Technologia de los Alimentos
      Santiago, Chile
      Steven Zeisel MD PhD
      University of North Carolina
      Chapel HilI, NC, USA
Observers
      Elio Riboli MD ScM MPH
      Imperial College London, UK
WCRF Executive
      Marilyn Gentry
      President WCRF Global Network
      Kate Allen
      Director, WCRF International
      Deirdre McGinley-Gieser
      Senior Vice President for Programs, AICR
                                                            20
</pre>

====================================================================== Einde pagina 37 =================================================================

<br><br>====================================================================== Pagina 38 ======================================================================

<pre>Secretariat
      Martin Wiseman FRCP FRCPath (chair)
      Medical and Scientific Adviser
      WCRF International
      Rachel Thompson PhD PHNutr
      Science Program me Manager (Nutrition)
      WCRF International
      Susan Higginbotham PhD
      Director for Research, AICR
Imperial College London
      Teresa Norat PhD
      Principal Investigator, Continuous Update Project
      Imperial College London
      Doris Chan
      Research Associate, Continuous Update Project
      Imperial College London
      Rosa Lau
      Research Associate, Continuous Update Project
      Imperial College London
      Rui Veira
      Data Manager, Continuous Update Project
      Imperial College London
                                                        21
</pre>

====================================================================== Einde pagina 38 =================================================================

<br><br>====================================================================== Pagina 39 ======================================================================

<pre>Ref eren ces
1.    Sainsbury JR, Anderson TJ and Morgan DA. ABC of breast diseases: breast
      cancer. BMJ 2000; 321: 745-50.
2.    Fentiman IS, Fourquet A and Hortobagyi GN. Male breast cancer. Lancet
      2006; 367: 595-604.
3.    World Health Organization. The global burden of disease: 2004 update.
      2008.
4.    Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, 2002. CA Cancer
      J Clin 2005; 55: 74-108.
5.    Parkin DM, Whelan SL, Ferlay J, et al. Cancer Incidence in Five Continents,
      Vol. 1 to VIII. Lyon: IARC 2005. International Agency for Research on
      Cancer. Globocan 2002. http//wwwdep.iarcfrJ. 2006.
6.    International Agency for Research on Cancer. Globocan 2002. http:www
      dep.iarc.fr/. 2006.
7.    Jemal A, Clegg LX, Ward E, et al. Annual report to the nation on the status
      of cancer, 1975-2001, with a special feature regarding survival. Cancer
      2004; 101: 3-27.
8.    McPherson K, Steel CM and Dixon JM. ABC of breast diseases. Breast
      cancer epidemiology, risk factors, and genetics. BMJ 2000; 321: 624-8.
9.    Ries L, Eisner M, Kosary C, et al. SEER Cancer Statistics Review, 1975-
      2002. Bethesda, MD: National Cancer Institute. 2005.
10.   Kufe D, Pollock R, Weichselbaum R, et al. Holland Frei Cancer Medicine. 6
      ed. Hamilton, Ontario: BC Decker. 2003.
11.   Coleman MP, Quaresma M, Berrino F, et al. Cancer survival in five
      continents: a worldwide population-based study (CONCORD). Lancet Oncol
      2008; 9: 730-56.
12.   Bosetti C, Bertuccio P, Levi F, et al. Cancer mortality in the European
      Union, 1970-2003, with a joinpoint analysis. Ann Oncol 2008; 19: 631-40.
13.   MacMahon B. General Motors Cancer Research Prizewinners Laureates
      Lectures. Charles S. Mott Prize. Reproduction and cancer of the breast.
      Cancer 1993; 71: 3 185-8.
14.   Lippman ME, Dickson RB, Gelmann EP, et al. Growth regulation of human
      breast carcinoma occurs through regulated growth factor secretion. J CelI
      Biochem 1987; 35: 1-16.
                                                                                 22
</pre>

====================================================================== Einde pagina 39 =================================================================

<br><br>====================================================================== Pagina 40 ======================================================================

<pre>15. Murray PA, Barrett-Lee P, Travers M, et al. The prognostic significance of
    transforming growth factors in human breast cancer. Bri Cancer 1993;
    67: 1408-12.
16. International Agency for Research on Cancer. World Health Report 2008
    Editors P.Boyle, 3 Levin. Lyon. 2008.
17. Modan B, Chetrit A, Alfandary E, et a). Increased risk of breast cancer after
    low dose irradiation. Lancet 1989; 1: 629-31.
18. Post-menopausal oestrogen therapy. IARC Monogr Eval Carcinog Risks
    Hum 72:399-530. 1999.
19. International Agency for Research on Cancer. Hormonal Contraception and
    Post-menopausal Hormonal Therapy. In: IARC Monogr Eval Carcinog Risks
    Hum no 72.
    http://monographs.iarc.fr/ENGjMonographs/vol72/volume72.pclf. 1999.
20. Visvanathan K, Crum RM, Strickland Pl, et al. Alcohol dehydrogenase
    genetic polymorphisms, low-to-moderate alcohol consumption, and risk of
    breast cancer. Alcohol Clin Exp Res. 2007; 31: 467-76.
21. Wirfalt E, Mattisson 1, Gullberg B, et al. Fat from different foods show
    diverging relations with breast cancer risk in postmenopausal women. Nutr
    Cancer 2005; 53: 135-43.
22. Vogel U, Christensen J, Nexo BA, et al. Peroxisome profilerator-activated
    receptorgamma2 Prol2Ala, interaction with alcohol intake and NSAID use,
    in relation to risk of breast cancer in a prospective study of Danes.
    Carcïnogenesis 2007; 28: 427-34.
23. Nielsen NR and Gronbaek M. Interactions between intakes of alcohol and
    postmenopausal hormones on risk of breast cancer. Inti Cancer 2008;
    122: 1109-13.
24. Zhang SM, Lee IM, Manson JE, et al. Alcohol consumption and breast
    cancer risk in the Women’s Health Study. Am i Epidemiol 2007; 165: 667-
    76.
25. Tjonneland A, Christensen J, Olsen A, et al. Alcohol intake and breast
    cancer risk: the European Prospective Investigation into Cancer and
     Nutrition (EPIC). CancerCauses Control 2007; 18:361-73.
26.  Ericson U, Sonestedt E, Gullberg B, et al. High folate intake is associated
    with lower breast cancer incidence in postmenopausal women in the
     Malmo Diet and Cancer cohort. Am J Clin Nutr 2007; 86: 434-43.
27. Stolzenberg-Solomon RZ, Chang SC, Leitzmann MF, et al. Folate intake,
    alcohol use, and postmenopausal breast cancer risk in the Prostate, Lung,
    Colorectal, and Ovarian Cancer Screening Trial. Am J Clin Nutr 2006; 83:
    895-904.
                                                                                 23
</pre>

====================================================================== Einde pagina 40 =================================================================

<br><br>====================================================================== Pagina 41 ======================================================================

<pre>28. Mellemkjaer L, Bigaard J, Tjonneland A, et al. Body composition and breast
    cancer in postmenopausal women: a Danish prospective cohort study.
    Obesity (Silver.Spring) 2006; 14: 1854-62.
29. Smith-Warner SA, Spiegelman D, Yaun SS, et al. Alcohol and breast cancer
    in women: a pooled analysis of cohort studies. JAMA 1998; 279: 535-40.
30. Hamajima N, Hirose K, Tajima K, et al. Alcohol, tobacco and breast
    cancer— collaborative reanalysis of individual data from 53
    epidemiological studies, including 58,515 women with breast cancer and
    95,067 women without the disease. BrJ Cancer 2002; 87: 1234-45.
31. Suzuki R, Orsini N, Mignone L, et al. Alcohol intake and risk of breast
    cancer defined by estrogen and progesterone receptor status--a meta
    analysis of epidemiological studies. Int J Cancer 2008; 122: 1832-41.
32. Boffetta P and Hashibe M. Alcohol and cancer. The Lancet Oncology 2006;
    7: 149-56.
33. Andrieu N, Goldgar DE, Easton DF, et al. Pregnancies, breast-feeding, and
    breast cancer risk in the International BRCA1/2 Carrier Cohort Study
    (IBCCS). J Nat! Cancer Inst 2006; 98: 535-44.
34. Iwasaki M, Otani T, lnoue M, et al. Body size and risk for breast cancer in
    relation to estrogen and progesterone receptor status in Japan. Ann
    Epidemiol 2007; 17: 304-12.
35. Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer
    and breastfeeding: collaborative reanalysis of individual data from 47
    epidemiological studies in 30 countries, inciuding 50302 women with
    breast cancer and 96973 women without the disease. Lancet 2002; 360:
    187-95.
36. Ma H, Bernstein L, Ross RK, et al. Hormone-related risk factors for breast
    cancer in women under age 50 years by estrogen and progesterone
    receptor status: resuits from a case-control and a case-case corn parison.
    Breast Cancer Res 2006; 8: R39.
37. Lahmann PH, Friedenreich C, Schuit AJ, et al. Physical activity and breast
    cancer risk: the European Prospective Investigation into Cancer and
     Nutrition. Cancer Epidemiol Biomarkers Prev 2007; 16: 36-42.
38. Tehard B, Friedenreich CM, Oppert JM, et al. Effect of physical activity on
    wornen at increased risk of breast cancer: results from the E3N cohort
    study. Cancer Epidemiol Biomarkers Prev 2006; 15: 57-64.
39. Chang SC, Ziegler RG, Dunn B, et al. Association of energy intake and
    energy balance with postmenopausal breast cancer in the prostate, lung,
    colorectal, and ovarian cancer screening trial. Cancer Epidemiol
    Bïomarkers Prev 2006; 15: 334-41.
                                                                                24
</pre>

====================================================================== Einde pagina 41 =================================================================

<br><br>====================================================================== Pagina 42 ======================================================================

<pre>40. Bardia A, Hartmann LC, Vachon CM, et al. Recreational physical activity
    and risk of postmenopausal breast cancer based on hormone receptor
    status. Arch Intern Med 2006; 166: 2478-83.
41. Lundqvist E, Kaprio J, Verkasalo PK, et al. Co-twin control and cohort
    analyses of body mass index and height in relation to breast, prostate,
    ovarian, corpus uteri, colon and rectal cancer among Swedish and Finnish
    twins. Int J Cancer 2007; 121: 810-8.
42. Palmer JR, ms-Campbell LL, Boggs DA, et al. A prospective study of body
    size and breast cancer in black women. Cancer Epidemiol Biomarkers Prev
    2007; 16: 1795-802.
43. Reeves GK, Pirie K, Beral V, et al. Cancer incidence and mortality in
    relation to body mass index in the Million Women Study: cohort study. BMJ
    2007; 335: 1134.
44. Reinier KS, Vacek PM and Geller BM. Risk factors for breast carcinoma in
    situ versus invasive breast cancer in a prospective study of pre- and post
    menopausal women. Breast Cancer Res Treat 2007; 103: 343-8.
45. Lukanova A, Bjor 0, Kaaks R, et al. Body mass index and cancer: resuits
    from the Northern Sweden Health and Disease Cohort. lnti Cancer 2006;
    118: 458-66.
46. Michels KB, Terry KL and Willett WC. Longitudinal study on the role of body
    size in premenopausal breast cancer. Arch Intern Med 2006; 166: 2395-
    402.
47. Ahn J, Schatzkin A, Lacey iv, Jr., et al. Adiposity, adult weight change, and
    postmenopausal breast cancer risk. Arch Intern Med 2007; 167: 2091-
    102.
48. Krebs EE, Taylor BC, Cauley JA, et al. Measures of adiposity and risk of
    breast cancer in older postmenopausal women. i Am Geriatr Soc 2006;
    54: 63-9.
49. Gallicchio L, McSorley MA, Newschaffer Ci, et al. Body mass,
    polymorphisms in obesity-related genes, and the risk of developing breast
    cancer among women with benign breast disease. Cancer Detect Prev
    2007; 31: 95-101.
50. Suzuki R, Rylander-Rudqvist T, Ye W, et al. Body weight and
    postmenopausal breast cancer risk defined by estrogen and progesterone
    receptor status among Swedish women: A prospective cohort study. lntJ
    Cancer 2006; 119: 1683-9.
51. van den Brandt PA, Spiegelman D, Yaun SS, et al. Pooled analysis of
    prospective cohort studies on height, weight, and breast cancer risk. Am i
    Epidemiol 2000; 152: 514-27.
                                                                                 25
</pre>

====================================================================== Einde pagina 42 =================================================================

<br><br>====================================================================== Pagina 43 ======================================================================

<pre>52. Renehan AG and al. e. Body-mass index and incidence of cancer: a
    systematic review and meta-analysis of prospective observational studies.
    Lancet 2008; 371: 569-78.
53. Key T, Appieby P, Barnes 1, et al. Endogenous sex hormones and breast
    cancer in postmenopausal women: reanalysis of nine prospective studies.
    J NatI Cancer Inst 2002; 94: 606-16.
54. Campagnoli C, Abba C, Ambroggio S, et al. Pregnancy, progesterone and
    progestins in relation to breast cancer risk. The Journal of Steroid
    Biochemistry and Molecular Biology 2005; 97: 441-50.
55. Hilakivi-Clarke L, Forsen T, Eriksson JG, et al. Tallness and overweight
    during childhood have opposing effects on breast cancer risk. BrJ Cancer
    2001; 85: 1680-4.
56. Pasquali R, Pelusi C, Genghini S, et al. Obesity and reproductive disorders
    in women. Human reproduction update 2003; 9: 359-72.
57. Baer Hi, Rich-Edwards JW, Colditz GA, et al. Adult height, age at attained
    height, and incidence of breast cancer in premenopausal women. lnti
    Cancer 2006; 119: 223 1-5.
58. Rinaldi S, Key Ii, Peeters PH, et al. Anthropometric measures, endogenous
    sex steroids and breast cancer risk in postmenopausal women: a study
    within the EPIC cohort. lnti Cancer 2006; 118: 2832-9.
59. Tehard B and Clavel-Chapelon F. Several anthropometric measurements
    and breast cancer risk: resuits of the E3N cohort study. Int J Obes (Lond)
    2006; 30: 156-63.
60. Eliassen AH, Colditz GA, Rosner B, et al. Adult weight change and risk of
    postmenopausal breast cancer. JAMA 2006; 296: 193-201.
61. Michels KB, Xue F, Terry KL, et al. Longitudinal study of birthweight and the
    incidence of breast cancer in adulthood. Carcinogenesis 2006; 27: 2464
    8.
62. Innes K, Byers T and Schymura M. Birth characteristics and subsequent
    risk for breast cancer in very young women. Am J Epidemiol 2000; 152:
    112 1-8.
63. Hilakivi-Clarke L. Mechanisms by which high maternal fat intake during
    pregnancy increases breast cancer risk in female rodent offspring. Breast
    CancerRes Treat 1997; 46: 199-214.
64. Lof M, Sandin S, Lagiou P, et al. Dietary fat and breast cancer risk in the
    Swedish women’s lifestyle and health cohort. BrJ Cancer 2007; 97: 1570-
    6.
                                                                                26
</pre>

====================================================================== Einde pagina 43 =================================================================

<br><br>====================================================================== Pagina 44 ======================================================================

<pre>65. Sonestedt E, Guliberg B and Wirfalt E. Both food habit change in the past
    and obesity status may influence the association between dietary factors
    and postmenopausal breast cancer. Public Health Nutr 2007; 10: 769-79.
66. Thiebaut AC, Kipnis V, Chang SC, et al. Dietary fat and postmenopausal
    invasive breast cancer in the National Institutes of Health-AARP Diet and
    Health Study cohort. J Nat! Cancer Inst 2007; 99: 451-62.
67. Kim EH, Willett WC, Colditz GA, et al. Nurses’ Health Study (NHS) Cohort
    1976-l996Dietary fat and risk of postmenopausal breast cancer in a 20-
    year follow-up. Am i Epidemio! 2006; 164: 990-7.
68. Smith-Warner SA, Spiegelman D, Adami HO, et al. Types of dietary fat and
    breast cancer: a pooled analysis of cohort studies. lnti Cancer 2001; 92:
    767-74.
69. Prentice RL, Caan B, Chlebowski RT, et al. Low-fat dietary pattern and risk
    of invasive breast cancer: the Women’s Health Initiative Randomized
    Controlled Dietary Modification Trial. JAMA 2006; 295: 629-42.
70. Kaaks R, Berrino F, Key T, et al. Serum sex steroids in premenopausal
    women and breast cancer risk within the European Prospective
    Investigation into Cancer and Nutrition EPIC). J Nat! Cancer Inst 2005; 97:
    755-65.
71. Wu AH, Pike MC and Stram DO. Metaanalysis: dietary fat intake, serum
    estrogen levels, and the risk of breast cancer. J Nat! Cancer Inst 1999; 91:
    529-34.
72. Qin L, Xu J, Wang P, et al. Soyfood intake in the prevention of breast
    cancer risk in women: a meta-analysis of observational epidemiological
    studies. J NutrSci V!taminol(Tokyo) 2006; 52: 428-36.
73. Trock B, Hilakivi L and Clarke R. Meta-analysis of soy intake and breast
    cancer risk. J Nat! Cancer Inst 2006; 98: 459-71.
74. Wu A, Yu M, Tseng C, et al. Epidemiology of soy exposures and breast
    cancer risk. BrJ Cancer 2008; 98: 9-14.
                                                                              27
</pre>

====================================================================== Einde pagina 44 =================================================================

<br><br>====================================================================== Pagina 45 ======================================================================

<pre>Appendix 1 Criteria for grading evidence
(Taken from Chapter 3 of the Second Expert Report)
This box lists the criteria finally agreed by the    Panel that were necessary to support
the judgements shown in the matrices. The            grades shown here are ‘convincing’,
‘probable’, ‘limited suggestive’, ‘limited
                      —                         — no conclusion’, and ‘substantial effect on
risk unlikely’. In effect, the criteria define these terms.
Convincing
These criteria are for evidence strong enough to support a judgement of a convincing
causal relationship, which justifies goals and recommendations designed to reduce
the incidence of cancer. A convincing relationship should be robust enough to be
highly unhikely to be modified in the foreseeable future as new evidence accumulates.
All of the following were generally required:
     •   Evidence from more than one study type.
     •   Evidence from at least two independent cohort studies.
     •   No substantial unexplained heterogeneity within or between study types or in
         different populations relating to the presence or absence of an association, or
         direction of effect.
     •   Good quality studies to exclude with confidence the possibility that the
         observed association results from random or systematic error, inciuding
         confounding, measurement error, and selection bias.
     •   Presence of a plausible biological gradient (dose response’) in the
         association. Such a gradient need not be linear or even in the same direction
         across the different levels of exposure, so long as this can be explained
         plausibly.
     •   Strong and plausible experimental evidence, either from human studies or
         relevant animal models, that typical human exposures can lead to relevant
         cancer outcomes.
Probable
These criteria are for evidence strong enough to support a judgement of a probable
causal relationship, which would generally justify goals and recommendations
designed to reduce the incidence of cancer.
All the following were generally required:
     •   Evidence from at least two independent cohort studies, or at least five case
         control studies.
     •   No substantial unexplained heterogeneity between or within study types in the
         presence or absence of an association, or direction of effect.
     •   Good quality studies to exclude with confidence the possibility that the
         observed association results from random or systematic error, including
         confounding, measurement error, and selection bias.
     •   Evidence for biological plausibility.
Limited     —  suggestive
These criteria are for evidence that is too limited to permit a probable or convincing
causal judgement, but where there is evidence suggestive of a direction of effect. The
                                                                                          28
</pre>

====================================================================== Einde pagina 45 =================================================================

<br><br>====================================================================== Pagina 46 ======================================================================

<pre>evidence may have methodological flaws, or be limited in amount, but shows a
generally consistent direction of effect. This almost always does not justify
 recommendations designed to reduce the incidence of cancer. Any exceptions to this
require special explicitjustification.
All the following were generally required:
     •   Evidence from at east two independent cohort studies or at least five case
         control studies.
     •   The direction of effect is generally consistent though some unexplained
         heterogeneity may be present.
     •   Evidence for biological plausibility.
Limited    —   no conciusion
 Evidence is so limited that no firm conclusion can be made. This category represents
an entry level, and is intended to allow any exposure for which there are sufficient
data to warrant Panel consideration, but where insufficient evidence exists to permit
a more definitive grading. This does not necessarily mean a limited quantity of
evidence. A body of evidence for a particular exposure might be graded limited      — no
conclusion’ for a number of reasons. The evidence might be limited by the amount of
evidence in terms of the number of studies available, by inconsistency of direction of
effect, by poor quality of studies (for example, lack of adjustment for known
confounders), or by any combination of these factors.
When an exposure is graded ‘limited         —  no conclusion’, this does not necessarily
indicate that the Panel has judged that there is evidence of no relationship. With
further good quality research, any exposure graded in this way might in the future be
shown to increase or decrease the risk of cancer. Where there is sufficient evidence
to give confidence that an exposure is unlikely to have an effect on cancer risk, this
exposure will be judged ‘substantial effect on risk unlikely’.
There are also many exposures for which there is such limited evidence that no
judgement is possible. In these cases, evidence is recorded in the full CUP SLRs on
the Diet and Cancer Report website (www.dietandcancerreport.org). However, such
evidence is usually not included in the summaries.
Substantial effect on risk unhikely
 Evidence is strong enough to support a judgement that a particular food, nutrition, or
physical activity exposure is unlikely to have a substantial causal relation to a cancer
outcome. The evidence should be robust enough to be unlikely to be modified in the
foreseeable future as new evidence accumulates.
All of the following were generally required:
     •   Evidence from more than one study type.
     •   Evidence from at least two independent cohort studies.
     •   Summary estimate of effect close to 1.0 for comparison of high versus low
         exposure categories.
     •   No substantial unexplained heterogeneity within or between study types or in
         different populations.
     •   Good quality studies to exclude, with confidence, the possibility that the
         absence of an observed association resuits from random or systematic error,
                                                                                      29
</pre>

====================================================================== Einde pagina 46 =================================================================

<br><br>====================================================================== Pagina 47 ======================================================================

<pre>          inciuding inadequate power, imprecision or error in exposure measurement,
          inadequate range of exposure, confounding, and selection bias.
      •   Absence of a demonstrable biological gradient (‘dose response’).
      •   Absence of strong and plausible experimental evidence, either from human
          studies or relevant animal models, that typical human exposures lead to
          relevant cancer outcomes.
Factors that might misleadingly imply an absence of effect inciude imprecision of the
exposure assessment, an insufficient range of exposure in the study population, and
inadequate statistical power. Defects in these and other study design attributes
might lead to a false conciusion of no effect.
The presence of a plausible, relevant biological mechanism does not necessarily rule
out a judgement of ‘substantial effect on risk unlikely’. But the presence of robust
evidence from appropriate animal models or in humans that a specific mechanism
exists, or that typical exposures can lead to cancer outcomes, argues against such a
judgement.
 Because of the uncertainty inherent in conciuding that an exposure has no effect on
risk, the criteria used to judge an exposure ‘substantial effect on risk unhikely’ are
roughly equivalent to the criteria used with at east a ‘probable’ level of confidence.
Conclusions of ‘substantial effect on risk unlikely’ with a lower confidence than this
would not be helpful, and could overlap with judgements of ‘limited       —  suggestive’ or
 ‘limited  — no conclusion’.
Special upgrading factors
These are factors that form part of the assessment of the evidence that, when
 present, can upgrade the judgement reached. So an exposure that might be deemed
a ‘limited   —  suggestive’ causal factor in the absence, say, of a biological gradient,
 might be upgraded to ‘probable’ in its presence. The application of these factors
 (listed below) requires judgement, and the way in which these judgements affect the
final conclusion in the matrix are stated.
      •   Presence of a plausible biological gradient (‘dose response’) in the
          association. Such a gradient need not be linear or even in the same direction
          across the different levels of exposure, so long as this can be explained
          plausibly.
      •   A particularly large summary effect size (an odds ratio or relative risk of 2.0 or
          more, depending on the unit of exposure) after appropriate control for
          confounders.
      •    Evidence from randomised trials in humans.
       •   Evidence from appropriately controlled experiments demonstrating one or
           more plausible and specific mechanisms actually operating in humans.
       •   Robust and reproducible evidence from experimental studies in appropriate
          animal models showing that typical human exposures can lead to relevant
          cancer outcomes.
                                                                                          30
</pre>

====================================================================== Einde pagina 47 =================================================================

<br><br>====================================================================== Pagina 48 ======================================================================

<pre>World Cancer American Institute

Research Fund for Cancer Research
International ne

www.werf.org www.aicr.org

Ki at
DE
hd
World Cancer
Research Fund

ww w.werf-uk.org

Wereld Kanker
Onderzoek Fonds

www.werf-nl.org

World Cancer
Research Fund
Hong Kong

LEN UAR AAE

www.werf-hk.org

Continuous
Update Project

www.dietandcancerreport.org C U Pp

©
©
=
t
=
c
(=)
<
”
Cc
a=)
2.
D
et
DO
=D
~
ol
D
0
tn
les)
3
D
o
n
or
ie)
&
=
(2)
D
=
No
Oo
a
O
AR
Ss
zl
3
t]
i
—

</pre>

====================================================================== Einde pagina 48 =================================================================

<br><br>====================================================================== Pagina 49 ======================================================================

<pre>                                                                                                                             IJc
                                                                                                                             International Journal of Cancer
Alcohol intake and breast cancer in the European prospective
investigation into cancer and nutrition
                                 , Véronique Chajès’, Jordi de Batlle, Carine Biessy’, Laure Dossus
Isabelle Romieu, Chiara Scoccianti
                                 1                                                                                       ,
                                                                                                                         4
                                                                                                                         3
                                                                                                                         ’
                                                                                                                         2
Laura Baglietto
          , Françoise Clavel-Chapelon
          6
          ’
          5                           , Kim Overvad
                                      4
                                      3
                                      ’
                                      2                               , Anja Olsen
                                                                      7                                   , Rudolf Kaaks
                                                                                          , Anne Tjønneland
                                                                                          8               8                        ,
                                                                                                                                   9
Annekatrin Lukanova’°, Heiner Boeing’    , Antonia ,
                                         1                     1314 Pagona ,
                                                               ’
                                                               12
                                                               Trichopoulou                        1415 Dimitrios Trichopoulos’
                                                                                                   ’
                                                                                                   13
                                                                                                   Lagiou                        ,
                                                                                                                                 5
                                                                                                                                 ”
                                                                                                                                 4
                                                                                                                                 ’
                                                                                                                                 2
Domenico Palli , Sabina Sieri
               16                 , Rosario Tumino
                                  17                 , Paolo ,
                                                     18                     20 Salvatore Panico
                                                                            ’
                                                                            19
                                                                            Vineis                    ,
                                                                                                      2t
H. B(as) ,   19              2 Carla H. van Gils
                             3
             Bueno-de-Mesquita
             2
             ’                                                 , Petra H. ,
                                                               24                  24 Eiliv Lund
                                                                                   ’
                                                                                   19
                                                                                   Peeters               , Gun Skeie
                                                                                                         25             ,
                                                                                                                        25
Elisabete ,   2526
              Weiderpass
              2
              ’            7 José Ramén Quirés Garcia
                           5                                      , Marfa-Dolores Chirlaque
                                                                  29                             , Eva ,
                                                                                                 3031                   32
                                                                                                                        ’
                                                                                                                        31
                                                                                                                        Ardanaz
Marfa-José ,   39
               Sénchez
               3
               ’          3 Eric J. Duell
                          4                   , Pilar Amiano
                                              35             , Signe Borgquist
                                                             36                    , Etisabet Wirfëlt
                                                                                   37                       , Göran Hallmans
                                                                                                            38                  ,
                                                                                                                                39
Ingegerd Johansson
             , Lena Maria Nilsson
             40                               , Kay-Tee Khaw
                                              41                    , Nick Wareham
                                                                    42               , Timothy J. Key
                                                                                     43                           , Ruth C. Travis
                                                                                                                  44                ,
                                                                                                                                    44
Neil Murphy
       , Petra A. Wark’
       19                       , Pietro Ferrari’ and Elio Riboli
                                9                                      19
  Nulritional Epiciemiology Group, Nutrition and Metabolism Sectien, International Agency for Research en Cancer, Lyon, France
  Inserm, Centre for Research in Epidemiology and Population Health (CESP), Uaoi8, Nulrilion. Hormones and Womens Health Team, Villejuif, France
  University Paris Sub, UMRS ioi8, Villejuit, France
lnstitut Gustave-Roussy, Villejuif, France
4
  Cancer Epidemiology Centre, Cancer Council Victoria, Melbourne, Australia
6 Centre for Epidemiology and Bioslatistics, The University of Melbourne, Melbourne, Australia
‘Department of Epidemiology, School of Public Health, Aar[ius University, Aarhus, Denmark
  Diet, Genes, and Environment Unit, Danish Canrer Society Research Center, Copenhagen, Denmark
  Division ei Caricer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany
   Departrnent of Medical Biosciences, Pathology, Umeg Iiniversity, Urne,i, Sweden
   Deparlmenl of Epidemiology, German Instilule of Human Nulrition, Potsdam-Rehbruecke, Nuthelal, Germany
   Hellenic Health foundalion, Athens, Greece
WHO Collaborstirig Center for Food and Nutrition Policies, Department of Hygiene, Epidemiology and Medical Statistics, University of Athene Medical
3
‘
   School, Alhens, Greece
   Bureau of Epidemiologic Research, Academy of Athens, Alhens, Greece
Key words: alcohol consumption, breast cancer, prospective study
Abbreviations: BC: breast cancer; BMI: body rnass index;; Cl: confidence interval; EPIC: European prospective investigation info cancer
and nutrition; ER: cstrogen receptor; FFQ: food-frequency questionnaire; FFl’P: first full-term pregnancy; HER2: human epidermal
growib factor receptor; ER: hazard ratio; PR: progesterone receptor
Additional Supporting Information may be found in the online version of this article.
This is an open accesa article under the tcrms of the Creative Comnsons Attribution-NonCornrnercial-NoDerivs I.icense, svhich perneits use
and distrihution in any medium, provided the original work is properly cited, the use is non-commercial and no modiflcations or
adaptations are made.
Grant sponsor: International Agency for Research on Cancer; Grant sponsors: European Commission (DG-SANCO; coordination of EPIC)
and the International Agency for Research on Cancer; Grant sponsor: Danish Cancer Society (Denmark); Grant sponsors: Ligue Contre le
Cancer, Institut Gustave Roussy, Mutuelle Générale de lEducation Nationale, and Institut National de la Santé et de la Recherche Médicale
(INSERM) (France); Grant sponsors: Deutsche Krebshilfe, Deutsches Krebsforschungszentrum and Fcderal Ministry of Education and
Research (Germany); Grant sponsors: Hellenic Health Foundation, the Stavros Niarchos Foundation, and the Hellenic Ministry of Health
and Social Solidarity (Grecce); Grant sponsors: Italian Association for Research on Cancer (AIRC) and National Research Council ([taly);
Grant sponsor: Dutch Ministry of Public Health, Welfare and Sports (VWS), Netherlands Cancer Registry (NKR), LK Research Funds,
Dutch Prevention Funds, Dutch ZON (Zorg Onderzoek Nederland); Grant sponsor: World Cancer Research Fund (WCRF) and Statistics
Netherlands (The Netherlands); Grant number: ERC-2009-AdG 232997; Grant sponsor: Nordforsk, Nordic Centre of Excellence
programme on Food, Nutrition and I-Iealth (Norwav); Grant sponsors: I-Teahh Research Fund (EIS), Regional Governrnents of Andalucfa,
Asturias, Basque Country, Murcia (project number 6236) and Navarra, lSClll RETIC (RDO6/0020) (Spain); Grant sponsors: Swedish
Cancer Society, Swedish Scientilic Counci] and Regional Government of Sk5ne and Vésterbotten (Sweden); Grant sponsor: Cancer Research
UK, Medical Research Council, Stroke Association, British I-Ieart Foundation, Departrnent of Health, Rood Standards Agency and Wellcome
Trust (United Kingdom)
DOl: l01002/ijc.29469
History: Received 21 Aug 2014; Accepted 3 Nov 2014; Online 9 Feb 2015
Correspondence to: Dr [sabelle Romieu, International Agency for Research on Cancer, 150 cours Albert Thomas, 69373 I.yon, Cedex,
France, Tel: +33472738094, Fax: +33472738361, E-mail: romieui@iarc.fr
Int. j. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. Published by Wiley Periodicals, lnc. on behalf of UICC
</pre>

====================================================================== Einde pagina 49 =================================================================

<br><br>====================================================================== Pagina 50 ======================================================================

<pre>1922                                                                                                                       Alcohol intake and breast cancer
“  Department of Epidemiology, Harvard School of Public Health, Boston, MA
 6
   Molecular and Nutritional Epidemiology Unit, Cancer Research and Prevention Institute ISPO, Florence, Italy
                                                                                           —
“ Epidemiology and Prevention Unit, Department of Preventive & Predictive Medicine, Fondazione IRCCS Istituto Na7ionale Dei Tumori, Milan, Italy
Cancer Registry and Histopathology Unit, Civic—MP.Arezzo” Hospital, ASP Ragusa, Italy
5                                             ‘,
   Departrnent of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, United Kingdom
   Unit of molecular and genetic epidemiology, Human Genetics Foundalion (HuGefl, Turin, Italy
   Department of Clinical and Experimenlal Medicinie Federico II Universily of Naples, Naples, Italy
   National Institute for Public HealIb and the Environment (RIVM), Bilthoven, The Netherlands
   Department of Gastroenterology and Hepatology University Medical Centre, Utrecht, The Netherlands
   Department of Epidemiology. lulius Center for F-lealth Sciences and Primary Care, University Medicaf Center Utrecht, Utrecht, The Netherlands
°‘ Department of Community Medicine, Faculty of Health Sciences, Urriversity of Tromsø, Tromsø, Norwuy
6
   Department of Genetic Epidemiology, Folkhdlsan Research Center, Helsinki, rinland
“Etiological Research Unit, Cancer Registry of Noiwny, Oslo, Norway
08
   Department of Medical Epidemiotogy and Biostatistics, Karolinska Inslitute, Stockholm, Sweden
   Public Health Directorate, Asturias, Spain
30
   Department of Epidemiology, Murcia Regional Healtfi Authority, Murcia, Spain
CIBER Epidemiology and Public Health CIBERESP, Madrid, Spain
3
   Navarre Public Health Institute, Pamplona, Spain
Aridalusian School of Public Health, Granada Bio-Health Research Institute (Granada.IBS), Granada, Spain
33
   Instituto De Investigaciön Biosanitaria De Granada, Granada, Spain
Unit of Nutrition and Cancer, Catalan Institute of Oncology (ICO-IDIBELL), Barcelona, Spain
3
36
   Public Health Division of Gipuzkoa, Gipuzkoa, Spain
“Division of Oncology, Department of Clinical Sciences, Lund University, Lund, Sweden
   Department of Clirncal Sciences, Lund University, Malmö, Sweden
   Department of Public Heulth and Ctinical Medicine, Nutritional Research, Umed University, Umed, Sweden
°
   Depurtment of Odontology, timeS University, Umed, Sweden
‘
   Public Heulth and Clinicul Medicine/Nutritional Research, UmeS University, Umeg, Sweden
   University of Cambridge, School of Clinical Medicine, Clinical Gerontology Unit, Cambridge, United Kingdom
“ MRC Epictemiology Unit University of Cambridge Schoot of Clinirat Medicine, Cambridge, United Kingdom
   Cancer Epidemiology Unit, Nuflield Depurtment of Medicine, University of Oxford, Oxford, United Kingdom
    Alcohol intake has been associated to breast cancer in pre and postmenopausal women; however results are inconclusive regard
    ing tumor hormonal receptor status, and potential modifying factors like age at start drinking. Therefore, we investigated the rela
    tion between alcohol intake and the risk of breast cancer using prospective observational data from the European Prospective
    Investigation into Cancer and Nutrition (EPIC). Up to 334,850 women, aged 35—70 years at baseline, were recruited in ten Euro
    pean countries and followed up an average of 11 years. Alcohol intake at baseline and average lifetime alcohol intake were calcu
    lated from country-specific dietary and lifestyle questionnaires. The study outcomes were the Hazard ratios (HR) of developing
    breast cancer according to hormonal receptor status. During 3,670,439 person-years, 11,576 incident breast cancer cases were
    diagnosed. Alcohol intake was significantly related to breast cancer risk, for each 10 g/day increase in alcohol intake the HR
    increased by 4.2% (95% Cl: 2.7—5.8%). Taking 0 to 5 g/day as reference, alcohol intake of>5 to 15 g/day was related to a 5.9%
    increase in breast cancer risk (95% Cl: 1—11%). Significant increasing trends were observed between alcohol intake and ER+/
    PR+, ER’—/PR—-, HER2— and ER—/PR—HER2— tumors. Breast cancer risk was stronger among women who started drinking prior
    to first full-time pregnancy. Overall, our results confirm the association between alcohol intake and both hormone receptor posi
    tive and hormone receptor negative breast tumors, suggesting that timing of exposure to alcohol drinking may affect the risk.
    Therefore, women should be advised to control their alcohol consumption.
    What’s new?
    Although it is now established that alcohol consumption increases breast cancer risk, many questions remain. Using a pro
    spective study design with 11,576 incident breast cancer cases across 10 European countries, the authors confirmed the
    increased risk of alcohol on breast cancer development. They further show that women who started drinking before their first
    full-term pregnancy have a higher risk than women who started afterwards. These effects were observed in hormone-receptor
    positive and —negative tumors pointing to non-hormonal pathways that need to be further investigated,
                                 Int. 1. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. Published by Wiley Periodicals, mc. on behatf of UICC
</pre>

====================================================================== Einde pagina 50 =================================================================

<br><br>====================================================================== Pagina 51 ======================================================================

<pre>Romieu et al.                                                                                                                    1923
A consistent association bas been observed between alcohol            IARC ethical committec and the local ethical committees of
intake and breast cancer (BC) among both pre and post                 the participating centers.
menopausal women) with a linear dose-response increase
ranging from 2%2 to 12% for each additional drink per day             Dietary assessment, lifestyle and alcohol consumption
(equivalent to about 10 g/day). While the association is firmly       Dietary and lifestyle questionnaires were compieted by partic
established, some questions such as the association with spe          ipants at enrolment when anthropon2etric rneasurernents
cific tumor sublypes, the impact of the age at start drinking                3 Past-year physical activity (PA) in occupational
                                                                      were taken)
and a potential window of susceptibility, remain unanswered.          and recreational domains was assessed at baseline with a self
Mechanistic evidences show that ethanol stimulates both ccli          administered questionnaire. For occupational activity, both
proliferation and estrogen receptor (ER) signaling in the             employment status as well as the level of physical activity
mammary gland. 6 Ivlost epidemiological studies report an             done during vork was recorded as: nonworker, sedentary,
                               7 1-lowever a recent meta
impact of ethanol on ER+ tumors.                                      standing, manual, heavy manual and unknown (for which
analysis showed an increased risk in both hormone receptor            duration and frequencies were not recorded). Recreational
                          8 The consumption of alcoholic
positive and negative tumors.                                         time physical activity included walking, cycling and sport
beverages may interact with other BC risk factors such as             activities. The duration and frequency of recreational activity
hormonal status or first full-term pregnancy (FFTP),
                                                   and
                                                   913                were multiplied by the intensity assigned by metabolic equiv
thus differentially modulate breast cancer risk over a vom            alent values (METs) for the different activities. A total PA
an’s lifetime» Recent studies report that low to moderate             index, the “Cambridge PA Index” was estimated by crossta
alcohol intake between menarche and first pregnancy is asso           bulating occupational with recreational PA. This index is
ciated with BC risk) 2 It is, therefore, important to evaluate        based on occupational, cycling and sport activities.
the association of alcohol intake and BC phenotypes in light              Information on alcohol use at the time of enrolment into
of a potential modulating effect of age at start drinking.            the study was based on a dietary assessn1ent of usual consun1p-
                                                                      tion of alcoholic beverages and types of alcoholic beverage (ie.,
                                                                      wine, beer, spirits and liquors) during the past 12 months. In
MateriaL and Methods                                                  each country, intake was calculated based on the estimated
The European Prospective Envestigation into Cancer and                average glass volume and ethanol content for each type of alco
Nutrition (EPIC) cohort consists of approxirnately 370,000            holic beverage, using information collected in highly standar
women and 150,000 men, aged 35—69, recruited hetween                  dized 24-hr dietary recalis from a subset of the cohort.N
1992 and 1998 in 23 research centers across 10 Western                Information on past alcohol consumption (available for 75.9%
European countries, Denmark (Aarhus and Copenhagen),                  of participants) was assessed as glasses of different beverages
France, Germany (1-leidelherg and Potsdam), Greece, Italy             consumed per week at 20, 30, 40 and 50 years of age. Average
(Florence, Varese, Ragusa, Turin, and Naples), Norway, Spain          lifetime alcohol intake was determined as a weighted average
(Asturias, Granada, Murcia, Navarra, and San Sehastian),              of intalce at differeni ages, with weights equal to the time of
Swedcn (Malrnö and Umeâ), the Netherlands (l3ilthoven and             individual exposure to alcohol at different ages. To determine
Utrecht) and the United Kingdom (Cambridge and Oxford).               which women had started drinking prior to FFTP, we used
The design and inethodology bas been published elsewhere.
                                                      13              information on alcohol consumption at different ages and the
Eligible men and women 3’ere invited to participate; those            age of PUT P reported by the women in the questionnaire.
who accepted gave informed consent and compiled question
naires on diet, lifestyle, and medical history. EPIC recruited        Anthropometric measurements
367,993 wornen, aged 35—70 years. Women with prevalent                Weight and height were measured al baseline, while the sub
cancers al any site at recruitment (n 19,853) or with miss-           jects were not wearing shoes. to the nearest 0.1 kg, or to the
ing diagnosis or censoring date (n 2,892) were excluded. A            nearest 0.1, 0.5 or 1.0 cm, depending on the center.’3 In
total of 3,339 subjccts with missing dietary or lifestyle infor       France, Norway and Oxford, height and weight were self
mation, and 6,753 wornen in the top and bottom 1% of the              reported on a questionnaire. The procedures used to account
ratio of energy intake to estimated energy requirement, calcu         for procedural differences between centers in the collection of
lated from age, sex, body weight and height, were excluded            anthropometric measurements are described elsewhere.’
                                                                                                                     6
from the analysis. In addition, 217 nonfirst breast cancer
cases were exciuded. Thus, the analysis was performed in               Perspective ascertainment of breast cancer cases, coding
334,850 EPIC women with complete exposure information.                of receptor status and determination of menopausaL status
Within this group, 11,576 wornen with invasive breast cancer           Incident BC cases were identified through population cancer
 (including 1,227 carcinoma in situ) were identified after a           registries (Denmark, Italy, the Netherlands, Norway, Spain,
median follow-up of t 1.0 years. Information on lifetime alco         Sweden and United Kingdom) or by active follow-up (France,
hol consurnption was missing for Sweden, Norway, Naples               Germany, Naples and Greece). The active follow-up proce
and Bilthoven, 24.1% were then excluded from the subanaly             dure used a combination of methods, including health insur
ses on lifetime alcohol intake. The study was approved by             ance records, cancer and pathology registries and contacts
 Int. 1. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. PubLished by Wiley Periodicals, mc. on behaif of UICC
</pre>

====================================================================== Einde pagina 51 =================================================================

<br><br>====================================================================== Pagina 52 ======================================================================

<pre>1924                                                                                                    Alcohol intake and breast cancer
with participants and their next-of-kin. Suhjects were fol           included in the models: age at menarche (<12, 12—14, >14
lowed up from study entry and until cancer diagnosis (except         years), age at birth of first child (nulliparous, <21, 21—30,
for nonmelanoma skin cancer cases), death and ernigration            >30 years), and age at menopause (<50, >50 years), ever
or until the end of the follow-up period, whichever occurred         use of contraceptive pill and ever use of replacement hor
first. The end of follow-up period was: December 2004 (Astu          niones, energy intake without alcohol consumption and
rias), December 2006 (Florence, Varese, Ragusa, Granada and          adjustment for interaction “menopause, weight.”
San Sehastian), December 2007 (Murcia, Navarra, Oxford,                  Alcohol consumption was niodeled as both continuous
Bilthoven, Utrecht and Denmark), June 2008 (Cambridge),              and categorical variable (none, 0.1—5, 5.1—15, 15.1—30, >30 g/
December 2008 (Turin, Malmo, Umea and Norway). For                   day). Both baseline consurnption and lifetinne consumption
study centers with active follow-up, the last follow-up contact      were studied. Correlation hetween both estinnations was high
was: December 2006 for France, December 2009 for Greece,             (r = 0.80). P-trend values were ohtained by modeling a score
June 2010 for Fleidelberg, December 2008 for Potsdam and             variable (from 1 to 5) category-specific level of alcohol at
December 2006 for Naples. Cancer incidence data were das             baseline. In addition, the shape of the dose-response curve
sified according to the International Classification of Diseases     between alcohol consurnption and breast cancer risk was
for Oncology, Second Revision (ICDO-2).                              evaluated with fractional polynomials of order two,   25 using
    Information on tumor receptor status, on the available lab       3 g/day as reference value and after exclusion of former con
oratory methods and on quantification descriptions used to           sunners at baseline. Nonlinearity was tested comparing the
determine receptor status, were collected by 20 centers. Infor       difference in log-likelihood of a model with fractional poly
mation on ER, progesterone receptor (PR) and human epider            nornials with a model with a linear term only to a chi-square
mal growth factor receptor 2 (HER2) was provided by each             distribution with three degrees of freedom.
                                                                                                           25 For all models,
center based on pathology reports. To standardize the quanti         the proportional hazards assumption was satisfied, evaluated
fication of receptor status among the EPIC centers, the follow       Via inciusion into the disease niodel of interaction terms
ing criteria for a positive receptor status were used: 10%           between exposure and attained age (data not shown). Statisti
celis stained, any “plus-system” description, 20 fmol/mg, an         cal heterogeneity of associations across countries or receptor
                                                      17— 1
AlIred score of >3, an IRS >2 or an FI-score >10.         -          status, was based on a 2 statistics, computed comparing
    Women were considered as prernenopausal when report              country-specific coefficients to an overall coefficient. Stratified
ing regular menses over the past 12 months, or when aged             analyses were conducted according to the time at start drink
<46 years at recruitment. Wornen were considered as post             ing (prior of after FFTP) and interaction term was tested
menopausal when not reporting any menses over the past 12            using alcohol intake as continuous variable in multivariate
months, or having received bilateral ovariectomy. Women              niodels. Models were run with the exclusion of the first 2
with missing or incomplete questionnaire data or with previ          years of follow-up, but the results did not differ from those
ous hysterectomy, vere considered postmenopausal only if             including the entire cohort (data not shown).
older than 55 years of age. Wonnen were considered with                   Statistical tests were two sided, and p-values <0.05 were
unknown menopausal status when aged between 46 and 55                considered significant. All analyses were perfornied using
years and had missing or incomplete questionnaire data, or           SAS version 9.2 (SAS Institute, 1999) and STATA (Stata Sta
reported previous hysterectomy (without ovariectomy).
                                               23
                                               ’
                                               22                    tistical Software: Release 12 (2011) StataCorp.,College Station,
                                                                     TX: StataCorp EP).
Statistical analysis
Cox proportional hazards regression models were used to              ResuLts
quantify the association between alcohol consumption and             During an average of 11.0 years of follow-up (3,670,43940
breast cancer risk. Age was the primary time variable and the        person-years) of 334,850 study participants, the EPIC study
Breslow method was adopted for handling ties.      1 ‘1ime at
                                                   ’
                                                   2                 documented 11,576 incident BC cases (e-1’ab1e 1). The over
entry was age at recruitnnent; time at exit was age at cancer        all percentage of women drinking over 15 g/day at baseline
diagnosis, death, loss to follow-up, or end of follow-up,            was 16.3% (e-Table 1).
whichever canne first. Models were stratified by center to con            The mean age at recruitment was 50.8 years, and the
trol for differences in questionnaire design, follow-up proce        mean age at BC diagnosis was 59.4 years. Table 1 presents
dures and other center effects. Further stratification by age at     the baseline alcohol intake according to the distribution of
recruitnient (l-year categories) was used. Systematic adjust         major baseline demographic and lifestyle characteristics. At
nnents were made for menopausal status (dichotonnized as             baseline, 35.2% of women were premenopausal and 43.1%
postmenopausal or women that underwent an ovariectomy                postmenopausal (the menopausal status of 18.8% of women
vs. other), weight and height (all continuous), smoking              was not defined, and 2.9% reported bilateral ovariectonny;
(never, former, and current), educational attainment (five cat       Table 1). No drinkers at baseline were less likely to ever have
egories of schooling) as a proxy variable for socioeconomic          used exogenous hormones and less likely to have ever
status, physical activity (inactive, moderately inactive, moder      smoked, were more moderately active and attained less edu
ately active, active). In addition, the following variables were     cation at baseline than drinkers at baseline (Table 1).
                          Int. j. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. Published by Wiley Periodicals, mc. on behalf of UICC
</pre>

====================================================================== Einde pagina 52 =================================================================

<br><br>====================================================================== Pagina 53 ======================================================================

<pre>   Table 1. Demographic and lifestyle characteristics according to breast cancer status and alcohol intake at baseline
                                                                                                                         Average daily alcohol intake (g/day)
    Demographic and lifestyle               Breast
    1
    characteristics                         cancer cases         Noncases              0                    0.1—5                5.1—15                15.1—30            >30
w
    Participants (N)                        11,576               323,274               54,907               135,599              89,694                35,460             19,190
    Age at recruitment (years) (mean, SD)   52.2 (8.8)           50.8 (10.2)           52.2 (9.0)           50.9 (9.8)           50.1 (9.5)            50.2 (9.0)         50.2 (9.0)
    Breast cancer cases (N, %)              11,576               —                     1,626 (2.96)         4,280 (3.16)         3,261 (3.64)          1,475 (4.16)       934 (4.87)
‘0
                        0/s)
    Receptor status (N,
    ER+/PR+                                 3,653 (31.6)                               527 (0.98)           1,367 (1.03)         970 (1.11)            472 (1.37)         317 (1.71)
    ER+/PR—                                 1,133 (9.8)                                177 (0.33)           404 (0.31)           303 (0.35)            162 (0.47)         87 (0.42)
    ER--/PR—                                1,050 (9.1)                                131 (0.25)           441 (0.33)           252 (0.29)            132 (0.39)         94 (0.51)
    HER—                                    1,764 (75.6)                               246 (0.46)           662 (0.50)           457 (0.53)            238 (0.70)         161 (0.81)
-1
    HER+                                    570 (24.4)                                 88 (0.16)            231 (0.18)           132 (0.15)            81 (0.24)          38 (0.21)
    ER—/PR—/HER—                            226 (2.0)                                  25 (0.05)            84 (0.06)            62 (0.07)             29 (0.09)          26 (0.14)
    Menopausal status (%)
       Premenopausal                        24.4                 35.2                  31.4                 37.2                 35.5                  32.5               29.1
       Perimenopausal                       22.0                 18.8                  17.3                 18.8                 19.2                  20.3               20.6
       Postmenopausal                       50.7                 43.1                  47.3                 41.4                 42.7                  44.5               47.1
      Surgical postmenopausal               2.8                  2.9                   4.0                  2.6                  2.6                   2.7                3.2
    Reproductive factors
      Age at menarche                       12.95 (1.51)          13.01 (1.75)         13.03 (1.57)         12.98 (1.68)         13.01 (1.65)          13.02 (1.55)       12.99 (1.56)
       (years) (mean, 50)
      Age at menopause                      49.52 (4.72)         49.03 (5.85)          48.80 (4.96)         49.09 (5.29)         49.20 (5.27)          49.17 (4.89)       49.07 (4.91)
       (years) (mean, SD)
                      2
       Nulliparous (%)                      13.7                  15.2                 10.8                 14.5                 17.3                  17.2               18.7
      N of full-term pregnancies            1.90 (1.17) (0—9)    1.99 (1.37) (0—17)    2.05 (1.21) (0—17)   2.01 (1.30) (0—14)   1.94 (1.28) (0—13)    1.90 (1.20) (0—12) 1.86 (1.20) (0—11)
       (mean, SD, range)
       Ever breastfed (%)                   71.7                 72.2                  75.3                 73.4                 71.1                  68.8               67.1
E   Exogenous hormone use (%)
       Ever-used HRT
                  2                         54.1                 42.2                  28.6                 42.4                 48.0                  48.8               50.9
       Ever-used CC                         58.8                 58.7                  40.7                 58.5                 65.2                  65.0               68.9
       Duration of CC use                   6.56 (6.95)          6.48 (9.15)           6.11 (7.08)          6.32 (8.18)          6.66 (8.06)           6.82 (7.27)        7.23 (7.21)
    Anthropometric factors
       Height (cm) (mean, SD)               161.62 (5.89)         161.08 (6.82)        160.48 (6.06)        160.94 (6.58)        161.42 (6.39)         161.70 (6.02)      162.02 (6.03)
      Weight (kg) (mean, 50)                66.20 (11.22)        65.70 (13.01)         66.60 (11.57)        65.96 (12.56)        64.99 (12.19)         64.88 (11.49)      65.38 (11.51)
                                                     Epidemiology
</pre>

====================================================================== Einde pagina 53 =================================================================

<br><br>====================================================================== Pagina 54 ======================================================================

<pre>                                                          Epidemiology
                                                                                                                                                                                                     “0
                                                                                                                                                                                                     0’
   Table 1. Demographic and lifestyle characteristics according to breast cancer status and alcohol intake at baseline (Continued)
                                                                                                                                   Average daily alcohol intake (g/day)
     Demographic and lifestyle                   Breast
     characteristics’                            cancer cases          Noncases              0                        0.1—5                  5.1—15              15.1—30              >30
       Waist-to-hip ratio (mean, SD)             0.80 (0.06)           0.80 (0.08)           0.80 (0.07)              0.80 (0.07)            0.79 (0.07)         0.80 (0.07)          0.80 (0.07)
                     3
        BMI (mean, SD)                           25.40 (4.08)          25.37 (4.73)          25.91 (4.20)             25.50 (4.56)           25.00 (4.42)        24.87 (4.17)         24.97 (4.18)
       Obese (BMI 30 kg/m  2
                           ) (%)                 11.4                  12.7                  22.3                     13.1                   8.9                 7.8                  8.1
     Smoking Status (%)
        Never smoker                             56.0                  57.0                  67.2                     58.8                   54.8                49.5                 39.2
        Former smoker                            25.0                  23.0                  14.8                     22.3                   26.3                27.5                 29.3
       Current smoker                            19.0                  20.0                  18.1                     19.0                   18.9                23.0                 31.5
     Total physical activity (%)
        nactive                                  17.8                  16.0                  9.0                      15.5                   20.0                20.1                 22.1
        Moderately inactive                      41.3                  37.1                  32.3                     37.3                   39.0                39.7                 41.8
        Moderately active                        34.4                  39.1                  51.4                     38.8                   33.6                33.2                 29.5
       Active                                    6.6                   7.8                   7.3                      8.4                    7.4                 7.0                  6.6
     Highest education level (%)
        None or primari! school                  26.1                  29.7                  52.5                     28.8                   22.1                21.8                 17.4
       Seconclary/technical/                     48.6                  46.8                  34.7                     50.1                   49.2                47.1                 49.5
        professional school
        University                               25.3                  23.5                  12.9                     21.2                   28.7                31.1                 33.1
0
     Dietary intake (mean, 50)
       Total energy intake (kcal/day)            1.976 (512)           1962 (594)            1,868 (522)              1,926 (567)            1,993 (550)         2,086 (519)          2206 (520)
       Total energy without                       1,918 (505)          1909 (586)            1,868 (521)              1,913 (566)            1,928 (549)         1,939 (518)          1902 (519)
        alcohol (kcal/day)
        Total dietaty fiber (g/day)               22.1 (7.1)           22.1 (8.2)            22.1 (7.3)               22.4 (8.0)             22.2 (7.7)          21.6 (7.3)           20.6 (7.3)
a  Note: Unknown values were excluded from the calculations. HRT: hormone replacement therapy; OC: oral contraceptives; SD: standard deviation; DM1: body mass index; All p values <0.0001, eecept
   for age at menarche (not signifficant); Trend test for continuous variables; Cochran—Arrnitage test for trend for categorical variables and global y’ test.                                       9.
8  Missing data in the total cohort were: 3.2°I for age at menarche; 4.6% for panty; 2.5% for oral contraceptive; 2.3% for smoking status; 15.3% for physical activity; 7.7% for diabetes; 14.4% for
   hypertension; 28.4% for waist-to-hip ratio; 3.9% for education level; in postmenopausal women: 5.3% for HRT and 24.3% for age at menopause.
   ‘Continuous variables are presented as means and standard deviations (SD), adlusted by age at recruitment and center (except age, which is adjusted by center only).
   Among postmenopausal wornen only.
   2
   3
   weight (kg)/height (m)’.
cz
</pre>

====================================================================== Einde pagina 54 =================================================================

<br><br>====================================================================== Pagina 55 ======================================================================

<pre> Romieu et al.                                                                                                                   1927
     Alcohol intake showed a significarit positive dose-response     Discussion
 relation with BC (p <0.0001, Table 2). BC ha’zard ratio (KR)        In this prospective study of 334,850 women and 11,576 inci
was increased by 6% (95% CI: 1—11%), 12% (95% Cl: 6—                 dent BC cases, an increased intake of 10 g of alcohol/day was
 19%) and 25% (9596 Cl: 17—35%) for the consurnption of 5—           related to a 4.296 increased BC risk (9596 Cl: 2.7—5.896). This
 15 g/day, 15—30 g/day and >30 g/day, respectively, compared         was observed for both ER+/PR± and ER—IPR— tumor sub-
to the 0.1—5 g/day category of intake. For each additional           types with the largest risk observed for triple negative tumors
 10 g/day the HR increased by 496 (9596 Cl: 3—6%). Figure 1          (ER—/PR—/HER2—). No interaction was observed with BMI
shows the relation between alcohol intake and BC risk, frac          and use of hormones. Women who started drinking before
tional polynomial of order 2 using 3 g/day as refercnce. A           their FFTP appeared to be at higher risk for BC than women
statisticallv significant relation was observed (p < 0.000 1),       who started drinking after their FFTP.
while the test for nonlinearity was compatible with a linear             Most studies published to date have reported an increased
trend (p = 0.100).                                                   BC risk with increasing alcohol intake.1 A previous analysis
     When the associations were evaluated according to hor           within the EPIC cohort on a smaller number of BC cases
mone receptor status, for each additional 10 g/day the HR            (u = 4,285), reported a 396 increase in BC incidence for each
signilcantly increased by 496 (9596 Cl: 1—696) in ER+/PR+,           additional 10 g/day of alcohol.26 Our results, based on
by 5% (9596 CI: 0—1096) in ER—/PR—, by 5% (9596 CI: 2—               [mt] 11,000 incident BC cases, confirm our previous results
9%) in HER2— and by 12% (95% CI: 3—23%) in ER—/PR—/                  and suggest a slightly stronger association. We did not observe
HER2— breast tumors (Table 2). ‘fest for heterogeneity               strong differences in estimates across tumor receptor status
between alcohol consumption and horrnone receptor status             (triple negative turnors showed the strongest risk, however, the
was not significant (p = 0.26). No significant association was       sample size in this category was small). Although most of prior
observed for ER+/PR—, ER—/PR+ and HER2+. When                        studies have reported a higher risk for ER± and/or PR+
using lifetlrne alcohol intake slightly lower estimates were         tumors compared to ER— and/or PR— tumors in particular,
observed (see eTable2), Similar results were observed for pre                                                         27 33 an
                                                                     for the highest versus the lowest alcohol intake group,
                                                                                                                      ’
                                                                                                                      9
and postmenopausal women, although, given the smaller                increased risk for hormone receptor negative lumors was also
sample size among premenopausal women, statislically signif          8
                                                                     reported.
                                                                     3
                                                                     ’       1 This inconsistency of results across studies
                                                                             5
icance was reached only in the overall analysis. There was no        might be partially due to the smaller number of BC cases with
heterogeneity in results hetween pre and postmenopausal              negative hormone receptor status. The very large number of
women (p interaction = 0.48). No interaction was observed            both hormone receptor positive and hormone receptor nega
with body mass index (BMI) or use of exogenous hormones              tive tumors in our study increased our power on the associa
either. Since statistical adjustrnent for smoking can be diffi       tion. Nonhormonal pathways such as DNA damage are likely
cult, analyses in nonsmokers at baseline were carried out and        to be involved in the incidence of receptor negative tumors.
                                                                                                                             8
results remained virtually similar (data not shown).                 The effect of alcohol appears linear, suggesting that there is no
     Age at start drinking according to FFTP, was positively         safe level of intake for BC risk.
related to BC risk among women who start drinking prior to               A limited number of studies have investigated the pres
FFTP. Stronger associations were observed for ER—, PR—,              ence of a window of susceptibïlity to alcohol carcinogenesis
ER—/PR— and ER—/PR—/HER2— tumors (Table 3). In a                     in the breast. Some epidemiological studies suggest that
multivariable model, an increase of 10 g of alcohol/day was          drinking alcohol during adolescence or early adulthood has a
related to au 8% (9596 Cl: 2—1496) increased risk of ER—             strong impact on BC rislc.36 Results from the Nurses’ Health
tumors in wornen who start drinking prior to FFTP, while no          Study II show that low to moderate alcohol intake during
association could be detected arnong women who start drink           adolescence and early adulthood is dose-dependently associ
ing after FFTP (p for interaction = 0.047), and a 996 (9596 Cl:      ated with an increased risk of proliferative benign breast dis
2—16%) increased risk of ER—/PR-— tumors in women who                ease, which may lead to invasive BC later in life.  37 More
start drinking prior to FFi’P (p for interaction = 0,10). When       recent results support the effect of drinking alcohol hetween
using lifetime alcohol intake slightly lower estirnates were         menarche and FFTP on BC risk (RR 1.11 per 10 g/day
observed (see eTable3). We were not able to evaluate the             intake; 95% CI: 1.00—1.23) and on proliferative benign breast
amount of alcohol consumed prior to FFTP.                            disease (RR= 1.16 per 10 g/day intake; 9596 CI: 1—1.02)» In
     BC hazard ratios, with data stratifled according to the         addition, the association between drinking before FFTP and
median period between menarche and FFTP (13 years)                   development of breast neoplasia appeared to he stronger with
arnong women who start drinking prior to FFTP, was of                langer menarche to first pregnancy intervals. These results
5.6% (9596 CI: 2.6—8.896) among women with longer median             are consistent with the hypothesis that alcohol carcinogens
period and of 2.696 (9596 Cl: 1.0—6.2%) among their counter          may preferentiallv act during mammary development.
                                                                                                                  38 We
part. These data suggest that a longer time between menarche         ohserved a stronger effect of alcohol intake prior to FFTP,
and FFTP may modulate BC risk among women who start                  with a significant interaction for receptor negative tumors.
drinking prior to FFTP. However, the test for interaction was        Our findings suggest that starting drinking before FF’l’P might
not significant (p = 0.23) (data not shown).                         be a more sensitive period, even if we cannot exclude the
Int. 1. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. Published by Wiley Periodicals, lnc. on behalf of UICC
</pre>

====================================================================== Einde pagina 55 =================================================================

<br><br>====================================================================== Pagina 56 ======================================================================

<pre>                                                          Epidemiology
                                                                                                                                                                                                           1.
                                                                                                                                                                                                           ‘0
    Table 2. Breast cancer risk by hormonal subtypes according to alcohol consumption at baseline
                                                                  Average daily alcohol intake at baseline (g/day)
D
      N cases/                                                                                                                                             p values       HR (95% Cl)
      person-years              01                       0.1—5                     5.1—15                  15.1—30                 >30                     2
                                                                                                                                                           (trend)        per 10 g/day             3
                                                                                                                                                                                                   p-value
      All cases                 1,626/605,217            4,280/1,488,055           3,261/983,711           1,475/387,280           934/206.177                            11,576/3,670,440
1                               1.04 (0.98—1.10)         1.00 (rel)                1.06 (1.01—1.11)        1.12 (1.06—1.19)        1.25 (1.17—1.35)        <.001          1.04 (1.03—1.06)          <.001
       ER+/PR-4-                527/598,625              1,367/1,470,607           970/969,575            472/381,199               317/202,569                           3,653/3,622,575
t.J                             1.06 (0.95—1.18)         1.00 (rel)                1.09 (1.00—1.18)        1.11 (0.99—1.23)         1.30 (1.15—1.48)       0.001          1.04 (1.01—1.06)         0.003
       ER-t-/PR—                177/596,367              404/1,464,086             303/965,046             162/379,089              87/200,949                            1,133/3,605,537
‘0
t))
0                               1.18 (0.98—1.42)         1.00 (rel)                1.13 (0.97—1.31)        1.22 (1.01—1.47)         1.13 (0.88—1.43)       0.41           1.04 (0.99—1.09)         0.09
0                               27/595,288               88/1,461,768              53/963.248              35/378,194               14/200,458                            217/3,598,956
t),
       ER—/PR+                  0.93 (0.59—1.45)         1.00 (rel)                1.06 (0.74—1.51)        1.37 (0.9ô—2.07)         1.03 (0.57—1.86)       0.26           1.05 (0.95—1.17)         0.34
0      ER—/PR—                  131/596,019              441/1,464.103             252/964,537             132/378,867             94/200,964                             1,050/3,604,490
‘S,
                                0.89 (0.73—1.10)         1.00 (rel)                0.92 (0.78—1.08)        1.03 (0.84—1.26)         1.28 (1.01—1.61)       0.06           1.05 (1.00—1.10)         0.03
S
(p                                                                                                         238/379,950             161/201.697                            1,164/3,612,130
       HER2—                    246/597,134              662/1,466,665             457/966,684
0
S.                              1.11 (0.95—1.29)         1.00 (rel)                1.09 (0.96—1.23)        1.14 (0.98—1.34)         1.41 (1.17—1.68)       0.007          1.05 (1.02—1.09)         0.004
0
vi     HER2+                    88/595,882               231/1,463,254             132/964,054            81/378,655                38/200,712                            570/3,602,557
0
0                               1.14 (0.87—1.49)         1.00 (rel)                0.89 (0.72—1.12)        1.18 (0.90—1.54)         0.97 (0.68—1.39)       0.83           0.98 (0.92—1.06)         0.68
0
S      ER—/PR—/HER2—            25/595,363               84/1,461,973              62/963,514              29/378,261               26/200,594                            226/3,599,705
(0
0.
0
                                1.09 (0.68—1.74)         1.00 (rel)                1.18 (0.84—1.66)        1.20 (0.77—1.86)         1.97 (1.23—3.16)       0.03           1.12 (1.03—1.23)         0.01
    l1nctudes both never and [armer drinkers.
-    Test for a trend in HRs by categories of alcohol intake were computed by assigning consecutive scores (1, 2, 3, 4, 5) to the categories.
     2
     3
     Tet of trend for alcohol intake continuous.                                                                                                                                                           >
    Note: Stratifled by center and age at recruitnient (1-year interval), and adjusted for menopausal status (pre/peri vs.postmenopausal women), oral contraceptive use (yes/no/missing), hormone          0
                                                                                                                                                                                                           S
9   replacement therapy use (yes/no/rriissing), height (continuous), weight (continuous), interaction menopause and weight, smoking status (never, ex, current and missing), educational level (primary,   0
St. no schooling, technical or professional or secondary, langer education, missing), physical activity (inactive, moderately active, moderately nactive, active and unknown), age at first menses (< 12.
     13—14. 15+, missing), age at first full term pregnancy (nulliparaus, <21, 22—30, >30, missing), age at menopause (<50, 50. missing) and energy intake without alcohol intake.
S                                                                                                                                                                                                          7’-
                                                                                                                                                                                                           (0
0
S                                                                                                                                                                                                          0.
                                                                                                                                                                                                           0
55
S                                                                                                                                                                                                          (0
0J                                                                                                                                                                                                         0)
0                                                                                                                                                                                                          n
                                                                                                                                                                                                           ci
5=                                                                                                                                                                                                         ci
                                                                                                                                                                                                           (0
</pre>

====================================================================== Einde pagina 56 =================================================================

<br><br>====================================================================== Pagina 57 ======================================================================

<pre> Romieu et al.                                                                                                                                         1929
 possibility that the stronger association between alcohol intake                 status could have differed from women with unavailable sta
 and BC in women who started drinking before FFTP might be                        tus. However, we did not observe such differences among
 the consequence of longer duration and amount of drinking.                       cases with known and unknown ER status and sub analyses
     In oor study, demographic characteristic, lifestyle and                      of these groups led to similar overall results. Similar strategies
 alcohol intake of women with available hormone receptor                          were adopted to inspect BC cases with and without available
                                                                                  information on PR and HER.2 status. In addition, a bias due
                                  Breast cancer
                                                                                  to the influence of preclinical disease on alcohol intake is
                          number of BC cases = 11,017                             unlikely, given that similar results were obtained after exclu
                                                                                  sion of samples from the first 2 years of follow-up. 1-lowever,
       12
                                                                                  we conducted multiple comparison analyses based on hormo
                                                                                  nal status and chance findings cannot be excluded.
       12
                                                                                      Major strengths of our study include the prospective and
       Ii                                                                         population based design, the large sample size, detailed infor
   HR                                                                             mation on alcohol intake at different period of life, age at
                                                                                  start drinking and types of beverage, data on hormone recep
                                                                                  tor status, excellent follow-up and large number of cases,
           L
           r--       r         —
                                             20
                                                                                  which provided us with good power for subgroups analyses.
                                                                                  Information on alcohol intake was self-reported and potential
                                                                                  misclassiflcation may have underestimated the effect of alco
          II               10                                  30                 hol intake. Still, assessment of alcohol intake has been shown
                        Alcohol irdake al recruilrcenl lawaai
                                                                                  to be reliable in the EPIC cohort
                                                                                                                  39m and the prospective set
Figure 1. Dose-response curve of BC risk with alcohol intake at
             —
                                                                                  ting of our study minimizes recall bias on age at start drink
recruitment. The dose-response curve is displayed up to 35 g/day,
                                                                                  ing and lifetime alcohol intake. We were unable to determine
corresponding to the 99th percentile of the alcohol intake distribu
tion. IColor figure can be viewed in tho online issue, which is avail             the amount of alcohol consumed befure f P and while
able at wileyonlinelibrary.com.]                                                  consumption both at baseline and over lifetime was
Table 3. Breast cancer risk among parous women with alcohol intake at baseline by age at start drinking before/after first full-term
pregnancy
                                                                                    Average daily alcohol intake at baseline
                                  Age at start                                                                                                Interaction
                                  drinking                   N cases/person-years        HR (95% Cl) for 10 g/day           p-value           p-value
                                                                                                                                              1
  All casea                       Before FFTP                4,104/1,216,204                 1.04 (1.02—1.06)                <.001            0.14
                                  After FFTP                 2,747/793,546                   1.02 (0.99—1.05)               0.26
  ER-l-                           Before FFTP                2,221/1,205,111                 1.04 (1.01—1.01)               0.005             0.16
                                  After FFTP                 1,460/786,197                   1.02 (0.98—1.01)               0.32
  PR+                             Before FFTP                1,315/1,199,890                 1.04 (0.99—1.01)               0.06              0.40
                                  After FFTP                987/783,211                      1.01 (0.96—1.07)               0.60
  ER-f/PR+                        Before WIP                 1,286/1,199,505                 1.04 (1.00—1.08)               0.04              0.39
                                  After FFTP                924/782.918                      1.01 (0.96—1.07)               0.65
  ER—                             Before FFTP                552/1,194,218                   1.08 (1.02—1.14)               0.009             0.05
                                  After FFTP                 371/778,873                     0.97 (0.88—1.06)               0.49
  PR—                             Before FFTP                776/1,196,034                   1.06 (1.02—1.11)               0.009             0.05
                                  After FFTP                 545/180,237                     0.98 (0.91—1.06)               0.66
  ER—/PR—                         Before FFTP               383/1,193,431                    1.09 (1.02—1.16)               0.01              0.10
                                  After FFTP                261/778,358                      0.97 (0.88—1.09)               0.65
  ER—/PR—/HER2—                   Before FFTP               99/1,191,822                     1.17 (1.04—1.31)               0.007             0.24
                                  After FFTP                 50/771,139                      0.97 (0.75—1.24)               0.78
t
A ge start prior to first full-term pregnancy (FFTP), was defined based on the information on ‘Age at start drinking alcohol’ and ‘Age at first full-term
pregnancy’. Results of stratified analyses hy age start prior/after FFTP are displayed. Signiftcance of interaction term was tested inciuding in a multi
variate model usieg alcohol as continuous variable and age start prior/after FFTP as categorical variable.
Note: Adlustments are the same as in Table 2. The statistical signiflcance of interactions was assessed using likelihood ratio tests based on the
models mith and without the interaction terms formed by the product of age at start drinking alcohol before or after first pregnancy and the value of
alcohol ietake at recruitment.
Int. J. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. Published by Wiley Periodicals, lnc. on behalf of UICC
</pre>

====================================================================== Einde pagina 57 =================================================================

<br><br>====================================================================== Pagina 58 ======================================================================

<pre>1930                                                                                                                               Alcohol intake and breast cancer
associated with a stronger adverse effect among women who                                 FFTP appears to have a larger adverse effect than after
start drinking prior to FFTP than among their counterpart,                                FTTP. No interaction with body fatness and use of hormone
our resuits should be interpreted with caution.                                           was observed. Alcohol has been shown to act through the
     In conciusion, findings from the EPIC cohort confirm the                             estrogen pathway, however our resuits suggest that nonhor
carcinogenic effect of alcohol intake on both receptor posi-                              monal pathways are likely to act and need to be further
tive and negative breast tumors. Starting to drink prior to                               investigated.
References
     IARC Monogr Eval Carcinog Risks Ilum.                 15. Friedenreich C, Cust A, Lahmann l’t-l, cl al.           type and hormone reccptor slatus in postmeno.
     Personal habits and indoor combustions. Volume            Anthropometric factors and risk of endometrial          pausal ivomen: the Nll-I-AARP diet and health
     100 E. A review of human carcinogens. !ARC                cancer: the european prospective investigation          study. Ani 1 Epideuniol 2009:170:308—17,
     Mouoçr Eval Carcinog Risks how 20l2;l00:t—                job cancer and nutrition. Cancer Causes Control     29. Kawase ‘t’, Matsuo K, 1 liraki A, eI al. lnteraclion
     538.                                                      2007;18:399—4l3.                                        of the etïeco of alcohol drinking and polymor
2.   Allen NE, Beral V, Casabonne D, ei al. Moderate       16. llal’tenherger M, Lahmann PIl, Panico S, cl al.         phisnis in alcohol-metabolizing enzymes 0fl Ihe
     alcohol intake and cancer incidence in women.             Overweight, obesity and fat distribution in 50- to      risk of female breasl cancer in Japan. 1 Epideuniol
     1 Natl Ca,rccr lust 2009:101:296—305.                     6t’year-old parlicipants in the European prospec        2009:19:21-t —50.
3.   Ellison RC, Zhang Y, McLennan CE, ei al.                  live investigation mln cancer and nutrition         30. ‘ferry MII, Knight JA, Zablotska L, cl al. Alcohol
     Exploring the relation of alcohol consumption to          (El’IC). Public llealth Nutr 2002:5:1147—62.            metabolism, alcohol intake, and brcast cancc’r
     risk of breast cancer. Am J Epidemiol 2001:154:       17. Harvey JM, Clark GM, Osborne CK, eI al. Estro           risk. a sistcr-set analysis using the breast cancer
     710—7.                                                    gen receptor status by inunanohistochemistry is         family registry. Breasi Cancer Res Treot 2007:106:
4.   Reichman ME, Judd j’r, Longcope C. eI al. Etlects         superior to the ligand-binding assay for predict        281-8.
     of alcohol consuEnption on plasma and urinary             ittg response to adjuvant endocrine therapy in      31. Chlebowski R’I’, Aoderson GL, Lane DS, cl al.
     hormone concentrations in preinenopausal                  breast cancer. 1 Clin Oicol 1999:17:1471—81.            Predidling risk of breasl cancer in postnteno
     women. 1 NoU Canccr lust 1993:85:722—7.               18. tstcCann J. Better assoys needed for hormone
                                                               ,                                                       pausal womcn by hormone receplor status. 1 NatI
5.   Pan S, Meng Q, Gao B, et al, Alcohol stimulates           receptor status, experts say. 1 NatI Caurer lust        Cancer lust 201(7:99:1695—705,
     estrogen receptor signaling in human breast can           2001:93:579—80.                                     32. Seliawan VW, Monroe KR, Wilkens IR, cl al.
     cer cr11 lines. Caucer Rc.s 2000:60:5635—9.           19. l.ayfield LJ, Gupta D, Mmmey EE. Assessment ol          Itreast cancer risk factors defined by estrogen and
6.   Ileldring N, Pike A, Andersson S, ei al. Estrogen         tissue estrogen and progesterone rcceptor levels: a     progcsteronc recc’ptor status: the multiethnic
     receptors: how do Lhey signal and what are their          survcy of current practice, techniques, and quan.       cohort study. /tun 1 Epideisiiol 2009:169:1251—9.
     targets. Ph
              ’siol Eer 210)7:87:905—31.
              3                                                titation methods. Brmst 1 2000;6:t89—96,            33. Li Cl, Chlcbowski 01’, Freiberg bi, cl al. Alcohol
7.   Coronado GD, Beasley J, Livaudais J. Alcohol         2)1. l:lowers IL, Burton GV, Cox Pit, cl al. Use ol          consumption and risk of poslmenopausal breast
     consumption and risk of breast cancer.Salud               monoclonal antieslrogen receptor antibody to eval       cancer by subtype: the women’s health initialive
     Publica Me.s 2011:93:110—7.                               uale estrogen receplor content in fine needie aspi.     observational sludy. 1 Nat? Caucer buit 2(110:102:
8.   Suzuki R, Orsini N, Mignone L, ei al. Alcohol             ralion breast biopsies. dun Surg 1986:203:250—1.        1422-31.
     iutake and risk of breast caocer defined by estro’   21. Remmele SV, Stegncr IIE. Recommendation ftir         34. Barnes 110, Steiodorf K, 1 1cm R, cl al. Populalion
     gen and progesterone receptor status—a meta               uniform definition of an imtnunoreactive score          altrihutable risk of invasive poslinenopausal
     analysis of epidemiological studies. Int 1 Caucer         (105) for imrnunohistochemical estrogen receptor        hreast cancer and breast cancer sublypes for
     2008;122:1832—’tl.                                        detection (ERICA) in breast cancer tissue,              niodifiable and non’inodifiahle risk factors. Con-
9.   Chen WY, Rosner 0, ilankinson SE, cl al. Mod’             Pathology 1987;8:t38-40.                                eer Epiderniol 2(111:35:315—52.
     erate alcohol consumption during adult life,         22. Kaaks 11, Rinaldi S, Key ‘f1, ei al. Postmenopausal  35. Kabal GC, Kim M, l’hipps Al, cl al. Smoking and
     drinking patterns, and breast cancer risk. JAMA           serum androgens, oestrogens and breast cancer           alcohol consuniption in relalion to risk of triple
     201 l;306:188’l—90.                                       risk: the european prospeclive investigation fl10       negalive breast cancer in a cohort of poslmeno.
10.  ‘t’erry MI), Zhang lF, Rabat G, eI al. Lifetime           cancer and nutrition. Endocr Relat Canser 2005:         pausal wonsen. Cancer Causes Control 201 t;22:
     alcohol intake and breast cancer risk. Anu Ëpide’           12:1071—82.                                           775—83.       -
     sniol 2006:16:230—40.                                 23. Kaaks R, iterrino F, Key T, cl al. Serum sex 51e-   36. Fcrnandez SV. Eslrogcn, alcohol consumption,
It.  Onland-Moret NC, Pee)ers Pil, van der Schouss’             roids in premenopausal wornen and breast cancer        and breast cancer. Alcohol Clin ExJi lle.s 2011:35:
     Y’l, eI al. Alcohol and endogenous sex steroid            risk within the european prospective invrstigation      389—9 1,
     levels in postmenopausal women: a cross                   into cancer and nutrition (EPIC). 1 NotI Cancer     37. Liu Y, Camimi EM, Berkey CS, cl al. tntakes of
     sectional study. 1 Clin Endocrinol Metab 2005;90:         lust 2005:97:755—65.                                    alcohol and folate during adolescence and risk of
     1111-9.                                               21. ‘t’hiehaut AC, Benichou 1. Choice of time-scale in      prohiferalive ben(gn breast disease. Pediolrics
12.  t.iu Y, Colditz GA, Rosner ii, cl al. Alcohol intake      cox’s model analysis of epidemiologic cohort data:      2012:129:1192—8.
     between menarche and first pregnancy: a pro’              o simulation stody. Stal Mcd 2001:23:3803-20.       38. Berkey CS, WilIclI WC, Frs’.ier Al., cl al. Pro
     spective study of breast cancer risk. 1 Natl Cancer   25. Sauerhrei W. Royston P. Building multivariable          spective sludy of adolescenl alcohol consumplion
     lust 2013:105:1571—8.                                     prognostic and diagnostic models: Iransformation        and risk of benign breast disease in young
13.  Riboli Ii, Hunt KJ, Slimani N, cl al. European            of the predictors by using fractiotsal polynomials.     wonsen. Pediatrks 2010:125:1081 —7.
     prospeclive investigation into cancer and nutri           JR Stalist Soc A 1999:162:71—9-t.                   39. ‘l’jonnel A, Overvad K, Flaraldsdottir J, cl al. Val
     tion (EPIC): study populations and data collec        26. ‘l’jonnel A, Christensen J, Olsen A, eI al. Alcohol     idalion of a semiquanlitalive food frequency
     tion. Pul,lic Iicalth Notr 21)02:5:1113—24.                intake and brcast cancer risk: Ihe european pro        qaeslionnaire developed in dcnmark. bul 1 Epide.
t’). Slimani N, Dehareeng G, Charrondiere RU, cl al.            spective investigation fl10 cancer and nutrition       mia? 1991:20:906—12.
     Structure of the standardized computerized 24 h            (EPIC). Cancer Causes Control 2007:18:361—73.      40. Kaaks R, Slimani NR. E pilol phasc studies on
     did recall interview used as reference method in      27. Zhang SM, Lee IM, Manson JE, eb al. Alcohol con         Ihe accuracy of dietary intake measurements in
     the 22 centers participating in the EPIC project.         sumplion and hreast cancer risk in the women’s          the El’IC project: overall evalaation of results.
     European prospective investiga)ion into cancer             health study. Arn 1 Epidemiol 2007;1 65:667—76.        European prospeclive investigation (fl10 cancer
     and nutrition. Comput Methods Programs Biorned        28. Lew JQ, l:rcedman ND, Leitzmann ME, cl al.              and nutrition. Int J Epidemiol 1997:26:36 Suppl 1:
     1999:58:251—66.                                            Alcohol and risk of breast cancer by histologic        S26—S36.
                                    Int. J. Cancer: 137, 1921—1930 (2015) © 2015 The Authors. Published by WiIey Periodicals, mc. on behalf of UICC
</pre>

====================================================================== Einde pagina 58 =================================================================

<br><br>====================================================================== Pagina 59 ======================================================================

<pre>                                                                                               824548
Van: Caroline van Rossum
Verzonden: maandag 24 augustus 2015 17:25
Aan: GR_RGV2O15
CC: Elly Buurma; Daphne van der A
Onderwerp: reactie vijfde ronde vanuit RIVM
Beste collega’s van de GR,
Hierbij de reactie vanuit het RIVM op de 5de ronde van de achtergronddocumenten RGV
Groetjes Caroline
Caroline van Rossum, PhD
Centre for Nutrition, Prevention and Health Services
National Institute for Public Health and the Environment
P0 Box 1
3720 BA Bilthoven
The Netherlands
See http://wwwvoedselconsumptiepeiIinQjj for information on the Dutch food consumption surveys
See http:/!wwwrivmnl/nevo for information on the Dutch food composition database
</pre>

====================================================================== Einde pagina 59 =================================================================

<br><br>====================================================================== Pagina 60 ======================================================================

<pre>Reactie RIVM op concept
achtergrondrapporten RGV ronde 5
dd 24-82015
Alcoholhoudende dranken.
   -  Pag 19: r 414: fysieke activiteit is een anglicisme/Belgisch, lichamelijke activiteit (zie regel
      421) is correcter Nederlands
   -  Pag 19, r 444: wijnconsumptie moet bierconsumptie zijn.
   -  Pag 29, r 690-691: de zin is incompleet.
   -  Pag 4, tabel 1: titel tabel 1: Gebruikelijke consumptie van ipv Gebruikelijke inname van..
                                                                   ...
Kalium
   -  Tabel 2 behoeft een update. Hier is correspondentie over geweest met Daphne van der A
      (10-7).
   -  Pagina 5, regel 69; VCP-2007-10 moet zijn VCP 2007-2010
   -  Komt in tabel 2 erbij wat voor gegevens de VCP gegevens zijn: Gebruikelijk?
Eiwit
   -  Tabel 1, correcte cijfers:
          o Jongens totaal eiwit: 11;13;16 ipv 13,15,18
          o Mannen totaal eiwit: 12;15;18 ipv 14,17,20
          o Meisjes : plant eiwit: p90 6 ipv 7.
          o Mannen: plan eiwit: plO: 4 ipv 5.
Visvetzuren
    - Regel 9, pagina 5: Volgens VCP 2011 moet zijn: Volgens VCP 2007-2010.
    - Regel 9: inhoudelijk tussen 2 en 19% is altijd lastig te interpreteren. Eventueel zou je ook
      kunnen verwijzen naar de aanvullende memo. Paragraaf 2.8 waarin we de supplementen
      voor de mannen en vrouwen rapporteren. Dan gebruik 10% van de mannen en 14% van de
      vrouwen visoliesupplementen.
</pre>

====================================================================== Einde pagina 60 =================================================================

<br><br>====================================================================== Pagina 61 ======================================================================

<pre>                                                                                     824541
Van: Peter de Wolf
Verzonden: maandag 24 augustus 2015 20:33
Aan: GR_Bibliotheek
CC: Weggemans, R.M. (Rianne)
Onderwerp: input STIVA op uw Concept Achtergronddocument Richtlijnen goede voeding 2015 -
Alcoholhoudende dranken
Geachte Gezondheidsraad,
Hierbij sturen wij u ons commentaar op uw Concept Achtergronddocument
Richtlijnen goede voeding 2015            - Alcoholhoudende dranken.
Er bestaan vele discrepanties tussen het eerder geschreven cEConcept
Achtergronddocument Richtlijnen goede voeding            -  1 en dit document
                                                            Alcohol
cEConcept Achtergronddocument Richtlijnen goede voeding            - Alcoholhoudende
. Dit zal het formuleren van voedingsadviezen ernstig bemoeilijken.
1
dranken
Verder menen wij dat de beschikbare wetenschappelijke gegevens slecht
bruikbaar zijn voor de gezondheidseffecten van specifieke alcoholhoudende
dranken. De gegevens zijn beperkt en bovendien van slechte kwaliteit er
kleven meerdere ernstige niethodologische tekortkomingen aan de
gerefereerde onderzoeken. Zo is in een aantal             - niet in het Concept
Achtergronddocument Richtlijnen goede voeding             - Alcoholhoudende dranken
                                                                            1
genoemde     -  wetenschappelijke publicaties aangetoond dat consumenten van
specifieke alcoholhoudende dranken ook verschillende leefstijlpatronen
vertonen (in het bijzonder voedingspatroon), die van invloed zijn op de
gevonden gezondheidseffecten. Dit maakt het lastig om gezondheidseffecten
toe te schrijven aan de consumptie van één specifieke alcoholhoudende
drank. Ook is het zo dat consumenten niet uitsluitend één specifieke
 alcoholhoudende drank drinken (en dus geen van de andere specifieke
 alcoholhoudende dranken).
Wij zien daarom geen meerwaarde in het werken aan Richtlijnen die
 verschillen per specifieke alcoholhoudende drank.
 Bijgevoegd ons uitgebreide en puntsgewijze commentaar.
 Met vriendelijke groet,
 Peter de Wolf
 Directeur
 1 $TV4]
 STIVA Stichting Verantwoorde AIcohoIconsumtie
 Parkstraat 15-25 1 2514 JD Den Haag 1
                      1 wiw.stiva.nl
  Volg ons op twitter         (&stivadewo!f
</pre>

====================================================================== Einde pagina 61 =================================================================

<br><br>====================================================================== Pagina 62 ======================================================================

<pre> Geachte Gezondheidsraad,
Hierbij sturen wij u ons commentaar op uw Concept Achtergronddocument Richtlijnen goede
voeding 2015 Alcoholhoudende dranken.
                —
Onze twee belangrijkste commentaren zijn:
Er bestaan vele discrepanties tussen het eerder geschreven ‘Concept Achtergronddocument
Richtlijnen goede voeding Alcohol’ en dit document ‘Concept Achtergronddocument Richtlijnen
                                —
goede voeding Alcoholhoudende dranken’. Dit zal het formuleren van voedingsadviezen ernstig
                  —
bemoeilijken.
Verder menen wij dat de beschikbare wetenschappelijke gegevens slecht bruikbaar zijn voor de
gezondheidseffecten van specifieke alcoholhoudende dranken. De gegevens zijn beperkt en
bovendien van slechte kwaliteit; er kleven meerdere ernstige methodologische tekortkomingen aan
de gerefereerde onderzoeken. Deze tekortkomingen worden niet overwogen.
Het ‘Concept Achtergronddocument Richtlijnen goede voeding 2015 Alcohol houdende dranken
                                                                         -
acht de bewijskracht voor een groot aantal verbanden ‘groot’ (paragrafen 3.8 en 4). We vinden
echter dat er onvoldoende wetenschappelijke basis is om drankspecifieke gezondheidseffecten te
kunnen benoemen.
In meer inhoudelijk detail is ons commentaar als volgt:
    1.   Er bestaat een groot aantal discrepanties tussen de conclusies getrokken in de
         achtergronddocumenten ‘alcohol’ en ‘alcoholhoudende dranken’. Daarom menen wij dat
         conclusies met betrekking tot dranktypen niet kunnen bijdragen aan een eventueel advies
         over de consumptie van specifieke dranktypen.
         Onduidelijk is dus hoe de verschillen tussen de uitkomsten gerapporteerd in het
         achtergronddocument alcoholhoudende dranken en de uitkomsten gerapporteerd in het
         achtergronddocument alcohol moeten worden geïnterpreteerd. Men zou immers een grote
         overeenkomst verwachten tussen de uitkomsten van de beide documenten. Dit is echter niet
         het geval; er zijn tegenstrijdigheden en veel informatie ontbreekt. De volgende
         tegenstrijdigheden vallen op in de conclusies van de beide documenten (zie ook Tabel 1):
             a. Het effect van alcohol op totale sterfte wordt alleen gevonden bij lagere doseringen
                   wijn; een tegengesteld effect op totale sterfte wordt gevonden bij lagere doseringen
                    bier en er is geen conclusie voor sterke drank
             b.     Het effect van alcohol op coronaire hartziekten wordt alleen gevonden voor wijn,
                   wordt gekenmerkt als ‘onwaarschijnlijk verband’ voor bier en er is geen conclusie
                   voor sterke drank
             c.    Over binge drinken is geen informatie beschikbaar
             d.     Over het effect van alcohol op beroerte kan niets worden geconcludeerd voor de
                    drie dranktypen, zowel bij lage als bij hogere consumptieniveaus
                                                        1
</pre>

====================================================================== Einde pagina 62 =================================================================

<br><br>====================================================================== Pagina 63 ======================================================================

<pre>            e.   Over het effect van alcohol op hartfalen kan niets worden geconcludeerd voor de
                 drie dranktypen
            f.   Het effect van alcohol op diabetes type II wordt alleen gevonden tot hoge doseringen
                 bij wijn, een tegengesteld effect op diabetes type II wordt gevonden bij bier
                 gedronken door mannen en lage consumptieniveaus van sterke drank gedronken
                 door mannen, bij vrouwen wordt een onwaarschijnlijk verband geconcludeerd
                 (terwijl voor alcohol juist een sterker verband lijkt te bestaan bij vrouwen)
             g.  Het effect van alcohol op darmkanker wordt min of meer vergelijkbaar
                 geconcludeerd voor bier en wijn, maar niet voor sterke drank (geen conclusie)
             h.  Het effect van alcohol op borstkanker is niet eenduidig voor alle drie de dranktypen
             i.  Over het effect van alcohol op dementie kan niets worden geconcludeerd voor de
                 drie dranktypen
Tabel 1: Overzicht conclusies met grote bewijskracht in de achtergronddocumenten ‘alcohol’ en
‘alcoholhoudende dranken’
                      Alcohol              Bier               Wijn                      Sterke drank
Totale sterfte         \I (< 30 g/d)       t (> 10 g/d M)     t(> 20 g/d V)             -
                                           t(>3g/dV)          t(>40g/dM)
                                                                    (< 10 g/d V)
                                                              ‘{, (< 20 g/d M)
Coronaire              1, (> 2,5 g/d)      0V                 j, (25 g/d)               -
hartziekten (CHZ)
CHZ binge              t                   -                  -                         -
drinken
Beroerte                   (< 15 g/d)      -                  -                         -
Beroerte               t (> 30 g/d)        -                  -                         -
Hartfalen              \I’ (< 28 g/d)      -                  -                         -
Diabetes type II       \I, (< 24 g/d V)     t (M)              ..j, (< 60 g/d)           t (< 12 g/d V)
                       \I,(<48g/dM)        OV(V)                                         OV(M)
Darmkanker             t (30-60 g/d)        t(20-40 g/d)       t(20-40 g/d)             -
 Borstkanker           t (> 10 g/d)         Niet eenduidig     Niet eenduidig            Niet eenduidig
 Dementie              {, ( < 30 g/d)      -                  -                          -
0V = onwaarschijnlijk verband
     2.  Een belangrijke methodologische kanttekening die wordt gemist is de correctie voor
         verstoring (confounding) in de vergelijking tussen de effecten van bier, wijn en gedistilleerde
         dranken. Een van de grote problemen bij het bestuderen van de effecten van de afzonderlijk
         alcoholhoudende dranken is dat de meeste consumenten zowel bier als wijn als gedistilleerd
         drinken; het komt zelden voor dat één dranktype uitsluitend wordt geconsumeerd. Veelal
         (ook in de studies vermeld in dit achtergronddocument) wordt een rekenkundige bewerking
         uitgevoerd om toch een effect van één specifieke dranksoort te kunnen afleiden. Een
         dergelijke bewerking houdt geen rekening met variaties in drinkpatronen (dagelijks wijn of
         wekelijks bier / voor of bij de maaltijd drinken) en variaties in andere factoren (geslacht;
         vrouwen drinken meestal wijn / leeftijd) en heeft dus tekortkomingen. Een directe
         vergelijking van de effecten van de drie dranktypen op de gezondheid uitsluitend door
                                                       2
</pre>

====================================================================== Einde pagina 63 =================================================================

<br><br>====================================================================== Pagina 64 ======================================================================

<pre>   middel van epidemiologisch onderzoek heeft dus grote methodologische nadelen en is dus
   niet verantwoord te maken.
3. Een tweede belangrijke methodologische kanttekening betreft de correctie van de overige
   leefstijifactoren (met name dieet) bij typische bierconsumenten, wijnconsumenten en
   consumenten van sterke drank. Een beroemd voorbeeld is de studie door Grønbaek’, die een
   duidelijk gezondheidsvoordeel liet zien voor de wijndrinker in vergelijking met de bierdrinker
   en de gedistilleerddrinker. Deze studie is later opnieuw geanalyseerd met een uitgebreidere
   correctie voor de voeding van de diverse typen drinkers
                                                      ; door deze correctie verdwenen de
                                                      2
   verschillen tussen bier, wijn en gedistilleerd helemaal. De rol van confounding in de relatie
   tussen dranktype en gezondheidsuitkomst is daarna nog eens door deze groep bevestigd.
                                                                                       3
   Het is dus zeer waarschijnlijk dat de wijndrinker een andere leefstiji (met name voeding)
   heeft dan de bierdrinker, waardoor de uitkomsten worden verstoord. Overigens, noemen
   Ferrari et al
               4 dit probleem ook in hun discussie: ‘Although we believe that this finding is
   relevant, we calI for cautious interpretations of these results, as the lifestyle profile of wine
   and beer drinkers is profoundly different.
4. Andere grote onderzoeken en reviews die geen effect van dranktype laten zien (op totale
   sterfte en coronaire hartziekten (paragrafen 3.2 en 3.3) worden niet mede overwogen in dit
   achtergronddocument. Deze onderzoeken zijn toegevoegd aan de referentielijst van dit
   commentaar.
   Mukamal et al  5 concluderen: “Among men, consumption of alcohol at least three to four
   days per week was inversely associated with the risk of myocardial infarction. Neither the
   type of beverage nor the proportion consumed with meals substantially altered this
   association. Men who increased their alcohol consumption by a moderate amount during
   follow-up had a decreased risk of myocardial infarction.”
   Rimm et al  6 concluderen in hun meta analyse: “Although most ecological studies support the
   hypothesis that wine consumption is most beneficial, the methodological problems of these
   studies limit their usefulness in drawing conclusions. Most of the differences in findings
    regarding specific drink types are probably due to differences in patterns of drinking specific
   types of alcoholic drink and to differing associations with other risk factors. Results from
   observational studies, where individual consumption can be assessed in detail and linked
   directly to coronary heart disease, provide strong evidence that a substantial proportion of
   the benefits of wine, beer, or spirits are attributable primarily to the alcohol content rather
   than to other components of each drink.”
    7 concludeert uit zijn systematische review: “1. Small doses of alcohol (1-4 drinks a
    Cleophas
    day) are associated with a slightly reduced risk of mortality and coronary heart disease
    (CHD). 2. Small doses (1-4 drinks a day) of wine, beer, and spirits are equally beneficial. 3.
    Apart from a direct beneficial effect of low doses of alcohol on mortality and CHD, some
    psychological factors may contribute to its beneficial effect.”
   Tolstrup en Gronbaek concluderen in hun review
                                                : Finally, there is some evidence that wine
                                                8
    may have more beneficial effects than beer and distilled spirits; however, these results are
    still controversial and may be confounded by personal characteristics and other lifestyle
    factors such as diet. The inverse association between alcohol intake and CHD is influenced by
    age, gender, drinking pattern, and possibly by type of alcohol.
                                                3
</pre>

====================================================================== Einde pagina 64 =================================================================

<br><br>====================================================================== Pagina 65 ======================================================================

<pre>   Klatsky et al
              9 concluderen: We conclude that (1) drinking ethyl alcohol apparently protects
   against coronary disease, and (2) there may be minor additional benefits associated with
   drinking both beer and wine, but not especially red wine...etc.
5. Het is te verwachten dat de conclusies getrokken in de achtergronddocumenten ‘alcohol’ en
   ‘alcoholhoudende dranken’ met betrekking tot totale sterfte en coronaire hartziekten
   (paragrafen 3.2 en 3.3) overeenkomen tussen alcohol en wijn, maar niet tussen alcohol en
   bier en gedistilleerd; er zijn immers meer studies uitgevoerd naar de effecten van
   wijnconsumptie dan dat er studies zijn uitgevoerd naar de effecten van bier- en
   gedistilleerdconsumptie.
   De conclusies in het achtergronddocument ‘alcoholhoudende dranken’ in de paragrafen 3.2
   en 3.3 zijn gebaseerd op een enkele meta analyse
                                               ° die een uitgebreidere versie is van een
                                               1
   eerdere meta analyse door grotendeels dezelfde groep epidemiologen. Door de uitbreiding
   van de meta analyse komen de auteurs tot een herziene conclusie. Costanzo et a1     °
                                                                                       1
   concluderen (zie abstract van): “In previous studies evaluating whether different alcoholic
                                10
   beverages would protect against cardiovascular disease, a J-shaped relationship for
   increasing wine consumption and vascular risk was found; however a similar association for
   beer or spirits could not be established. An updated meta-analysis on the relationship
   between wine, beer or spirit consumption and vascular events was performed              From 16
   studies, evidence confirms a i-shaped relationship between wine intake and vascular risk.
         Similarly, from 13 studies a J-shaped relationship was apparent for beer.(..). From 12
   studies reporting separate data on wine or beer consumption, two closely overlapping dose—
   response curves were obtained (maximal protection of 33% at 25 g/day of alcohol). This
   meta-analysis confirms the i-shaped association between wine consumption and vascular
   risk and provides, for the first time, evidence for a similar relationship between beer and
   vascular risk. In the meta analysis of 10 studies on spirit consumption and vascular risk, no J
   shaped relationship could be found.
    De auteurs melden in de discussie bovendien dat, data voor bier- en gedistilleerdconsumptie
    nog steeds beperkt zijn: “Unfortunately, the very limited data available about either beer or
    spirit consumption in relation to cardiovascular or total mortality, did not allow us to perform
    a fully meta-analytic investigation on the latter two beverages.”
    De conclusie geformuleerd door de auteurs is dus anders dan de conclusie weergegeven in
    het achtergronddocument (paragraaf 3.3.1). Deze laatste is gebaseerd op een andere
    analyse, een deelanalyse, uit hetzelfde artikel. Het is vooralsnog onduidelijk waarom het
    achtergronddocument deze analyse volgt en op basis van deze analyse de relatie tussen
    bierconsumptie en hart- en vaatziekten risico aanduidt als een onwaarschijnlijk verband en
    niet de uiteindelijke conclusie van de auteurs volgt.
    De in paragraaf 3.2 gerefereerde studie van Ferrari et al4 betreft met name wijndrinkers; +/-
     112.000 wijndrinkers tegenover +1- 31.000 bierdrinkers. Negentig procent van de vrouwen
    tegenover ongeveer vijftig procent van de mannen in deze studie worden gekenmerkt als
    wijndrinker.
                                                4
</pre>

====================================================================== Einde pagina 65 =================================================================

<br><br>====================================================================== Pagina 66 ======================================================================

<pre>6. Een derde belangrijke methodologische kanttekening wordt terecht gemaakt op pagina 7,
   namelijk dat er kritische opmerkingen zijn gemaakt over de controle groepen
   (geheelonthouder) in cohortonderzoeken (Fillmore et al’  ) naar de associatie tussen alcohol
                                                            2
   en ziekte uitkomsten. Het is echter voor de volledigheid goed te vermelden dat cohorten die
   wel een onderscheid hebben kunnen maken tussen niet-drinkers en ex-drinkers in hun
   controle groep, geen essentiële verschillen vonden in de beschreven associaties
                                                                          . Het ‘sick
                                                                          6
                                                                          ’
                                                                          13
   quitters’ argument lijkt dus niet te gelden. Ook wanneer de controle niet uit
   geheelonthouders bestaat maar uit lichte drinkers zijn er verdere dalingen van het risico
   .
   8
   ”
   7
   beschreven’
7. Met betrekking tot het interventieonderzoek, begrijpen we de keuze voor de intermediairen
   (bloeddruk, LDL cholesterol en BMI) zoals die wordt omschreven in het document ‘werkwijze
   van de commissie richtlijnen goede voeding 2015’. Wij betreuren de gekozen benadering
   echter in het geval van dit specifieke achtergronddocument.
   HDL cholesterol verhoging, c.q. HDL gemedieerde cholesterol efflux’
                                                                   9 en andere HDL functies
   worden niet meegewogen in het hoofdstuk 2: lnterventieonderzoek. Deze keuze is gemaakt
   omdat medicijnen en niacine die HDL cholesterol verhogen, niet aantoonbaar bijdragen aan
   het voorkomen van hartaanvallen. Er zijn echter een beperkt aantal geneesmiddelen getest
   dat HDL holesterol verhoogt, c.q. HDL functie verbetert en alcohol (net als lichamelijke
   activiteit) is een van de weinige nutriënten die niet alleen HDL cholesterol verhoogt maar ook
   zijn beschermende functies positief beïnvloedt’°”. HDL wordt in dezelfde mate verhoogd
   door bier, wijn en , 24 evenals de meeste andere intermediairen zoals
                        ’
                        22
                        gedistilleerd
   gerapporteerd in de meta-analyse van Brien’
                                            .
                                            4
   Bovendien wordt door het volgen van de cases geëvalueerd door het lOM’  5 een aantal
   andere belangrijke factoren die een causaal verband aannemelijk maken, zoals fibrinogeen
   en HbAlc niet geëvalueerd.
   Door deze benadering kan de commissie geen conclusie trekken over de effecten van
   alcoholhoudende dranken op geen enkele intermediair (zelfs niet LDL cholesterol, noch
   bloeddruk). Interventie onderzoek maakt echter zeer aannemelijk dat er een causaal verband
   is tussen consumptie van matige hoeveelheden alcoholhoudende dranken en een lagere
   incidentie van hart- en vaatziekten, zoals besproken in een systematisch review en meta
   4 en cohort studies’
   analyse’              .
                         6
8. In het werkwijze document wordt gesteld dat de commissie zich in beginsel beperkt in haar
   literatuuronderzoek tot een kritische evaluatie van gepoolde analyses, meta-analyses en
   systematische reviews die gepubliceerd zijn in peer-reviewed tijdschriften. In gepoolde
   analyses en meta-analyses worden de bevindingen uit meerdere oorspronkelijke
   onderzoeken met overeenkomstige vraagstelling en aanpak gecombineerd tot een nieuwe
   risicosch atti ng.
   Echter de conclusies met betrekking tot totale sterfte zijn gebaseerd op één multicenter
   , die wellicht voldoet aan het criterium ‘gepoolde analyse’, maar niet aan het
   4
   studie
   criterium ‘bevindingen uit meerdere oorspronkelijke onderzoeken gecombineerd tot een
   nieuwe risicoschatting’. Toch wordt de bewijskracht als ‘groot’ omschreven.
                                               5
</pre>

====================================================================== Einde pagina 66 =================================================================

<br><br>====================================================================== Pagina 67 ======================================================================

<pre>   9.   Het is opvallend dat met betrekking tot Diabetes Mellitus type 2 (paragraaf 3.4), op basis van
        het onderzoek van Beulens et al 27 het achtergronddocument conclusies trekt over de
        verschillende dranktypen, terwijl de auteurs conclusies trekken over ‘moderate alcohol
        consumption’ en niet over drank specifieke effecten. De auteurs merken in hun discussie op:
        “The specific risk reduction associated with wine consumption, however, appears to
        contradict the findings of several mechanistic studies. It was previously shown that the
        reduced risk of diabetes with moderate alcohol consumption can be explained by increased
        adiponectin concentrations for 25_30%28. However, randomized trials in study populations
        consuming a variety of alcoholic beverages could not detect a difference in the effects on
        adiponectin concentrations
                     . This suggests that the underlying biological mechanism is
                     2932
        most probably explained by alcohol itself. The specific risk reduction observed with wine
        could thus be attributed to other factors associated with wine consumption. Previous studies
        have shown that wine drinkers differ from drinkers of other beverages by consuming a
        healthier diet and being less likely to smoke
                                                . As men and women may also differ with
                                                33
        regard to such health-related behaviours, as is seen in the different structure of confounders
        amongst men and women, this could in part explain the specific association observed for
        wine consumption and the different effects between men and women.
    10. In paragraaf 3.4 wordt herhaaldelijk gerefereerd aan ‘aanvullend onderzoek’ van Cullmann
                                                                                            34
        en telkens wordt vermeld dat het onderzoek een te beperkt aantal cases betreft om daar
        conclusies op te baseren. Wellicht kan dit onderzoek worden verwijderd of minder worden
        benadrukt.
We hopen met bovenstaand commentaar een constructieve bijdrage te hebben geleverd aan het
Concept Achtergronddocument richtlijnen goede voeding 2015 Alcoholhoudende dranken.
                                                                  —
Met vriendelijke groet,
Peter de Wolf
Directeur STIVA
                                                    6
</pre>

====================================================================== Einde pagina 67 =================================================================

<br><br>====================================================================== Pagina 68 ======================================================================

<pre>Referenties
1      Gronbaek, M. et al. Mortality associated with moderate intakes of wine, beer, or spirits. BMJ
       (Clinical research ed.) 310, 1165-1169 (1995).
2      Tjonneland, A., Gronbaek, M., Stripp, C. & Overvad, K. Wine intake and diet in a random
       sample of 48763 Danish men and women. The American journa( of c!inical nutrition 69, 49-54
        (1999).
3       Gronbaek, M. Factors influencing the relation between alcohol and mortality--with focus on
       wine. J Intern Med 250, 291-308 (2001).
4       Ferrari, P. et al. Lifetime alcohol use and overall and cause-specific mortality in the European
        Prospective Investigation into Cancer and nutrition (EPIC) study. BMJ Open 4, e005245,
        doi:10.1136/bmjopen-2014-005245 (2014).
5       Mukamal, K. i. et al. Roles of drinking pattern and type of alcohol consumed in coronary
        heart disease in men. The New England journal of medicine 348, 109-118,
        doi:10.1056/NEJM0aO22O95 (2003).
6       Rimm, E. B., Klatsky, A., Grobbee, D. & Stampfer, M. i. Review of moderate alcohol
        consumption and reduced risk of coronary heart disease: is the effect due to beer, wine, or
        spirits. BMJ (Clinical research ed.) 312, 731-736 (1996).
7       Cleophas, T. J. Wine, beer and spirits and the risk of myocardial infarction: a systematic
        review. Biomed Pharmacother 53, 417-423, doi:10.1016/s0753-3322(99)80121-8 (1999).
8       Tolstrup, i. & Gronbaek, M. Alcohol and atherosclerosis: recent insights. Current
        atherosclerosis reports 9, 116-124 (2007).
9       Klatsky, A. L., Armstrong, M. A. & Friedman, G. D. Red wine, white wine, liquor, beer, and risk
        for coronary artery disease hospitalization. The American journal of cardiology 80, 416-420
        (1997).
10      Costanzo, S., Di Castelnuovo, A., Donati, M. B., lacoviello, L. & de Gaetano, G. Wine, beer or
        spirit drinking in relation to fatal and non-fatal cardiovascular events: a meta-analysis.
        European journal of epidemiology 26, 833-850, doi: 10.1007/s 10654-011-963 1-0 (2011).
11      Di Castelnuovo, A., Rotondo, S., lacoviello, L., Donati, M. B. & De Gaetano, G. Meta-analysis
        of wine and beer consumption in relation to vascular risk. Circulation 105, 2836-2844 (2002).
12      Filimore, K. M., Stockwell, T., Chikritzhs, T., Bostrom, A. & Kerr, W. Moderate alcohol use and
        reduced mortality risk: systematic error in prospective studies and new hypotheses. Ann
        Epidemiol 17, 516-23, doi:10.1016/j.annepidem.2007.01.005 (2007).
13      Mukamal, K. i., Chiuve, S. E. & Rimm, E. B. Alcohol consumption and risk for coronary heart
        disease in men with healthy lifestyles. Archives of internal medicine 166, 2145-2150,
        doi:10.1001/archinte.166.19.2145 (2006).
 14      Marmot, M. & Brunner, E. Alcohol and cardiovascular disease: the status of the U shaped
        curve. BMJ (Clinical research ed.) 303, 565-568 (1991).
 15      Ronksley, P. E., Brien, S. E., Turner, B. J., Mukamal, K. J. & Ghali, W. A. Association of alcohol
        consumption with selected cardiovascular disease outcomes: a systematic review and meta
         analysis. BMJ (Clinical research ed.) 342, d671, doi:10.1136/bmj.d671 (2011).
 16      Klatsky, A. L. & Udaltsova, N. Abounding confounding: sick quitters and healthy drinkers.
        Addiction (Abingdon, England) 108, 1549-1552, doi:10.1111/add.12157 (2013).
 17      Maclure, M. Demonstration of deductive meta-analysis: ethanol intake and risk of myocardial
         infarction. Epidemiol Rev 15, 328-351 (1993).
 18      Maclure, M. Alcohol intake and risk of myocardial infarction. The New England journal of
         medicine 330, 1241; author reply 1242 (1994).
 19      Rohatgi, A. et al. HDL cholesterol efflux capacity and incident cardiovascular events. The New
         Englandjournal of medicine 371, 2383-2393, doi:10.1056/NEJMoa14O9O65 (2014).
                                                        7
</pre>

====================================================================== Einde pagina 68 =================================================================

<br><br>====================================================================== Pagina 69 ======================================================================

<pre>20 Sierksma, A., van der Gaag, M. S., van Tol, A., James, R. W. & Hendriks, H. F. Kinetics of HDL
   cholesterol and paraoxonase activity in moderate alcohol consumers. Alcoholism, cIlnical and
   experimental research 26, 1430-1435, doi:10.1097/01.aic.0000030639.57507.60 (2002).
21 Sierksma, A. et al. Effect of moderate alcohol consumption on parameters of reverse
   cholesterol transport in postmenopausai women. Alcoholism, clinical and experimental
   research 28, 662-666 (2004).
22 van der Gaag, M. S. et al. Daily moderate alcohol consumption increases serum paraoxonase
   activity; a diet-controlied, randomised intervention study in middie-aged men.
   Atherosclerosis 147, 405-410 (1999).
23 van der Gaag, M. S. et al. Alcohol consumption stimulates early steps in reverse cholesterol
   transport. Journal of lipid research 42, 2077-2033 (2001).
24 Brien, S. E., Ronksley, P. E., Turner, B. J., Mukamal, K. i. & Ghaii, W. A. Effect of alcohol
   consumption on biological markers associated with risk of coronary heart disease: systematic
   review and meta-analysis of interventional studies. BMJ (Clinical research ed.) 342, d636,
   doi:10.1136/bmj.d636 (2011).
25 Micheel, C. M. & Bali, i. R. Evaiuation of Biomarkers and Surrogate Endpoints in Chronic
   Disease. Institute of Medicine Committee on Qualifications of Biomarkers and Surrogate
   Endpoints in Chronic Disease (2010).
26 Mukamai, K. J. et al. Drinking frequency, mediating biomarkers, and risk of myocardial
   infarction in women and men. Circulation 112, 1406-1413,
   doi:10.1161/circulationaha.105.537704 (2005).
27 Beuiens, J. W. J. et al. Alcohol consumptiori and risk of type 2 diabetes in European men and
   women: influence of beverage type and body size. The EPIC—InterAct study. Journal of
   internal medicine 272, 358-370 (2012).
28 Beulens, J. W., Rimm, E. B., Hu, F. B., Hendriks, H. F. & Mukamal, K. J. Alcohol consumption,
   rnediating biomarkers, and risk of type 2 diabetes among middle-aged women. Diabetes care
   31, 2050-2055, doi:10.2337/dcO8-0814 (2008).
29 Beuiens, J. W., de Zoete, E. C., Kok, F. i., Schaafsma, G. & Hendriks, H. F. Effect of moderate
   alcohol consumption on adipokines and insulin sensitivity in lean and overweight men: a diet
   intervention study. European journal of clinical nutrition 62, 1098-1105,
   doi:10.1038/sj.ejcn.1602821 (2008).
30 Beulens, J. W. et al. The effect of moderate alcohol consumption on adiponectin oligomers
   and muscle oxidative capacity: a human intervention study. Diabetologia 50, 1388-1392,
   doi:10.1007/s00125-007-0699-8 (2007).
31 Imhof, A., Plamper, t., Maier, S., Trischler, G. & Koenig, W. Effect of drinking on adiponectin in
   healthy men and women: a randomized intervention study of water, ethanol, red wine, and
   beer with or without alcohol. Diabetes care 32, 1101-1103, doi:10.2337/dcO8-1723 (2009).
32 Joosten, M. M., Witkamp, R. F. & Hendriks, H. F. Alterations in total and high-molecular
   weight adiponectin after 3 weeks of moderate alcohol consumption in premenopausal
   women. Metabolism: clinical and experimental 60, 1058-1063,
   doi:10.1016/j.metabol.2011.01.001 (2011).
33 Barefoot, i. C. et al. Alcoholic beverage preference, diet, and health habits in the UNC Alumni
   Heart Study. The American journal of clinical nutrition 76, 466-472 (2002).
34 Cullmann, M., Hilding, A. & Ostenson, C. G. Alcohol consumption and risk of pre-diabetes and
   type 2 diabetes development in a Swedish population. Diabetic medicine : a journal of the
   British Diabetic Association 29, 441-452, doi:10.1111/j.1464-5491.2011.03450.x (2012).
                                                  8
</pre>

====================================================================== Einde pagina 69 =================================================================

<br><br>====================================================================== Pagina 70 ======================================================================

<pre></pre>

====================================================================== Einde pagina 70 =================================================================

<br><br>====================================================================== Pagina 71 ======================================================================

<pre>                                                                                                      823831
Van: Lex Lemmers
Verzonden: vrijdag 21 augustus 2015 13:16
Aan: GR_RGV2O15
Onderwerp: Vijfde commentaarronde Richtlijnen goede voeding 2015 achtergronddocument
Alcoholhoudende dranken
Aan de gezondheidsraad,
Dank voor het bieden van de mogelijkheid om commentaar te leveren op het achtergronddocument
Alcoholhoudende dranken in het kader van de herziening van de Richtlijnen Goede Voeding. Namens
het Trimbos-instituut wil ik graag onderstaande punten bij u onder de aandacht brengen.
Als eerste zijn we verbaasd dat de relatie tussen alcohol en borstkanker als niet eenduidig wordt
gekenmerkt. We zijn benieuwd wat u precies bedoelt met niet eenduidig. Volgens onze lezing van de
literatuur is de relatie tussen borstkanker en alcoholconsumptie wel eenduidig en dit is onder andere
gebaseerd op bijgevoegde literatuur.
De vraag of gedistilleerd, bier of wijn van invloed is op de morbiditeit in het algemeen en kanker in
het bijzonder lijkt ons moeilijk te beantwoorden op basis van epidemiologisch onderzoek. Een
standaardglas gedistilleerd, bier of wijn bevat allemaal een zelfde hoeveelheid pure alcohol (10 gram
in Nederland) en hoewel verschillend qua concentratie alcohol in de drank, leiden ze tot een zelfde
BAC in het lichaam. De stof alcohol en het afbraakproduct van alcohol acetaldehyde worden beiden
als carcinogeen aangemerkt. Hoewel consumenten voorkeur kunnen hebben voor een specifieke
alcoholhoudende drank (wijn, bier of gedistilleerd), worden in de praktijk diverse alcoholhoudende
dranken door elkaar heen gedronken (bijvoorbeeld wijn in combinatie met een aperitief en cognac).
Met vriendelijke groet,
dr. Lex Lemmers
Wetenschappelijk medewerker B         - Jongeren en Riskant Gedrag
www.trimbos.nI
Da Costakade 45 3521 vs Utrecht
                    -
Postbus 725 3500 AS Utrecht
              -
Disclaimer
      Trimbos
      1
 I         instituut
 22 C            GGZ )Cennlsdag 2015
 2013            tit’.    zo,dIiqt4 andeeid
Denk aan het milieu voordat u deze e-mail print!
</pre>

====================================================================== Einde pagina 71 =================================================================

<br><br>====================================================================== Pagina 72 ======================================================================

<pre>                          British Journal of Cancer (2002) 87, 1234— 1245
            © 2002 Cancer Research UK All rights reserved 0007—0920/02 $25.00
                                                        www.bjcancer.com
Alcohol, tobacco and breast cancer                                                                —    collaborative reanalysis of
individual data from 53 epidemiological studies, including 585 1 5
women with breast cancer and 95 067 women without the
disease
       Collaborative Group on Hormonal Factors in Breast Cancer*’
       ‘Secretariat, Cancer Research UK Epiclemiology Unit Gibson Building, Radcliffe lnfsri-nary, Woodstock Road, Oxford 0X2 6Hft UK
      Alcohol and tobacco consumption are closely correlated and published results on their association with breast cancer have not
       always allowed adequately for confounding between these exposures. Over 80% of the relevant information worldwide on
       alcohol and tobacco consumption and breast cancer were collated, checked and analysed centrally. Analyses included 585 l 5
      women with invasive breast cancer and 95 067 controls from 53 studies. Relative hsks of breast cancer were estimated, after
       stratifying by study, age, panty and, where appropriate, women’s age when their first child was bom and consumption of alcohol
       and tobacco. The average consumption of alcohol reported by controls from developed countries was 6.0 g per day, ie. about
       half a unit/drink of alcohol per day, and was greater in ever-smokers than never-smokers, (8.4 g per day and 5.0 g per day,
       respectively). Compared with women who reported drinking no alcohol, the relative risk of breast cancer was 1.32 (1.19 1.45,              —
       P<0.00001) for an intake of 35—44 g per day alcohol, and 1.46 (1.33— 1.61, P<0.0000l) for ?45 g per day alcohol. The
       relative risk of breast cancer increased by 7.1% (95% Cl 5.5—8.7%; P<0.0000l) for each additional 10 g per day intake of
       alcohol, i.e. for each extra unit or drink of alcohol consumed on a daily basis. This increase was the same in ever-smokers and
       never-smokers (7.1% per 10 g per day, P<0.0000I, in each group). By contrast, the relationship between smoking and breast
       cancer was substantially confounded by the effect of alcohol. When analyses were restricted to 22255 women with breast
       cancer and 40832 controls who reported drinking no alcohol, smoking was not associated with breast cancer (compared to
       never-smokers, relative risk for ever-smokers 1.03, 95% Cl 0.98 1.07, and for current smokers0.99, 0.92— 1.05). The results
                                                                                      —
       for alcohol and for tobacco did not vary substantially across studies, study designs, or according to IS personal characteristics of
      the women; nor were the findings materially confounded by any of these factors. 1f the observed relationship for alcohol is
       causal, these results suggest that about 4% of the breast cancers in developed countnies are attributable to alcohol. In developing
       countries, where alcohol consumption among controls averaged only 0.4 g per day, alcohol would have a negligible effect on the
       incidence of breast cancer. In conclusion, smoking has little or no independent effect on the risk of developing breast cancer the
       effect of alcohol on breast cancer needs to be interpreted in the context of its beneficial effects, in moderation, on
       cardiovascular disease and its harmful effects on cirrhosis and cancers of the mouth, Iarynx, oesophagus and liver.
       British Journal oL Cancer (2002) 87, 1234— 1 245. doi: 10. 1 038/sj.bjc.6600596 wvvw.bjcancer.com
       © 2002 Cancer Research UK
       Keywords: breast cance alcohol; tobacco; smoking; collaborative reanalysis
Many epidemiological studies have investigated the relationship                         these exposures. Individual data from 65 epidemiological studies of
between breast cancer and the consumption of alcohol and/or tobac-                     breast cancer 63 published
                                                                                                            63 and two unpublished in which infor
co. References to over 80 studies that have collected relevant data, as                 mation on alcohol and/or tobacco consumption had been collected
well as to reviews of the subject, are given in Appendix II (www.                       contributed to this collaboration. These studies, some of which have
bjcancer.com). The published results from these studies have gener-                     not published resuits for alcohol or tobacco, include over 80% of the
ally suggested that women who regularly consume alcohol may be at a                     worldwide information on the topic (see Appendix II (www.bjcan
slightly increased risk of the disease, but the findings reported for                   cer.com)). The data from these studies were analysed, taking
tobacco are inconsistent. Alcohol and tobacco consumption are                           careful account of the possible confounding between alcohol and
known to be associated one with another, and published results have                     tobacco consumption, as well as confounding by other factors.
not always allowed adequately for possible confounding between
                                                                                        METHODS
*Correspondence. Valehe Beral, Secretariat, Cancer Research UK Epide
                                                                                        Eligibiity of studies and collection of data
miology Unit, Gibson Building, Radcliffe Infinmary, Oxfond 0X2 6HE. UK;
(See Appendix 1 for list of collaborators)                                              Data from epidemiological studies of women with breast cancer
Received 12 june 2002; revised 8 August 2002; accepted 23 August 2002                   have been brought together by the Collaborative Group on
</pre>

====================================================================== Einde pagina 72 =================================================================

<br><br>====================================================================== Pagina 73 ======================================================================

<pre>                                                                      Alcohol, tobacco and breast cancer
                                                                      Collaborative Group on Hormonal Factors in Breast Cancer
                                                                                                                                              235
Hormonal Factors in Breast Cancer to describe the relationship        tobacco consumption were stratified by alcohol consumption (0,
between breast cancer and various reproductive, hormonal and           <5, 5—14, 15—24, 25—34, 35—44, 45 g per day). In order to
other factors.
        6465 Case—control and cohort studies were eligible            summarise the relationship between alcohol consumption and
for the collaboration ifthey inciuded at least 100 women with inci    breast cancer risk, a linear trend in the log relative risk of breast
dent invasive breast cancer and recorded information on               cancer was fitted across increasing categories of consumption. In
reproductive factors and on use of hormonal therapies. Cohort         estimating such trends, the median consumption within a given
studies were inciuded using a nested case—control design, in which    category was taken to be the level of alcohol consumption for that
four controls were selected at random, matched on follow-up to        category.
the age of the casa at diagnosis and, where appropriate, broad            In general, results in the text are prasented as ralative risks and
geographical region. Data for individual women were collated          their appropniate 5E or FSEs. Where resuits are presented in the
and analysed centrally so that analyses could be carried Out using    form of plots, relative risks and their corresponding CIs/FCIs are
as similar definitions across studies as was possible. Details sought represented by squares and lines, respectively. The position of the
from principal investigators of each participating study included     square indicates the value of the relative risk and its area is inver
data that had been collected on each woman’s reproductive history     sely proportional to the vaniance of the logarithm of the relative
and various other factors that may be relevant to the aetiology of    risk, thereby providing an indication of the amount of statistical
breast cancer, inciuding the women’s consumption of alcohol and       information available for that particular estimata. Owing to the
tobacco.                                                              large number of relative risk estimates calculated, results are given
   Some investigators provided estimates of alcohol intake            with their appropriate 99% CIs/FCIs; and 95% CIs/FCIs are used
reported by each woman expressed as gram (g) of alcohol               only to summarise the main findings.
consumed per day or per week. Others provided information                 The absolute risk of breast cancer associated with various levels
                                                                                                                                               1
on the reported number of alcoholic drinks consumed daily or          of alcohol consumption was estimatad for woman in developad
weekly. In such instances, the number of grams of alcohol             countries, by applying the dose-response estimates obtained here
consumed per day, was estimated assuming that one alcoholic           to age-spacific incidence rates for breast cancer in daveloped coun
drink contains 12 g alcohol in the USA and Italy,’ t 8 g in the       tries around 199064,65 assuming that an intake of 10 g per day is
UK and 10 g elsewhere (Brewers’ Society, personal communica           roughly equivalent to one unit or drink of alcohol per day. The
tion). No information was sought about alcohol consumption at         cumulative incidence of breast cancer up to age 80 years was calcu
various ages or about the particular type of alcohol consumed.        lated from the age-specific findings.
Information was also sought on whether or not each woman
had ever smoked, and whether she was a current or past smoker.        RESULTS
Active smoking only was considered and no attention given to the
reported associations with environmental tobacco smoke,
                                                      49 nor
                                                      ’
                                                      35              The 65 studies that contributed individual data on alcohol and!
was information sought on the age women were when smoking             or tobacco consumption and other factors relevant to breast
started or stopped, or on the amount smoked. The methods of           cancer included a total of 66426 woman with invasive breast
identifying studies and of data checking, and correction, have        cancer (cases) and 126 953 woman without breast cancer controls
been described elsewhere.
                 65
                 ’
                 64                                                   from 63 published’          and two unpublished studies. Information
                                                                      on both alcohol and tobacco had been collected in 53 of these
Statistical analysis and presentation of resuits                      studies, that included a total of 58515 cases and 95067 controls
                                                                      from 51 published
                                                                                   ’ and two unpublished studies. Unless
                                                                                   5
Statistical methods ware similar to those used in previous reports    otherwise specified, analyses presented here are restricted to data
by this group.
          6467 Data from different studies were combined              from these 53 studies. This enables woman to be cross-classified
by means of the Mantel—Haenszel stratification technique, the         by both their alcohol and tobacco consumption, thus permitting
stratum-specific quantities calculated being the standard ‘observed   adequate examination of possible confounding between the two
minus expected’ (O—E) numbers of women with breast cancer,            exposures.
together with their variances and covariances. These values yield         Among woman with breast cancer in the 53 studies included in
both statistical descriptions (odds ratios, subsequently referred to  the main analyses, the median year of diagnosis was 1988 and the
as relative risks) and statistical tests (P values). When only two    average ae at diagnosis was 52.1 years. All but five of the 53
groups are being compared, relative risk estimates are obtained       ’” ware conducted in devalopad countries. Among
                                                                      9
                                                                      ’
                                                                      5
                                                                      studies
                                                                      4146
                                      66 as are their standard
from O—E values by the one-step method,                               controls, alcohol consumption was substantially greater in women
errors (SE) and confidence intervals (CI). When more than two         from developed than developing countries (average alcohol intakes
groups are compared, variances are estimated by treating the rela     of 6.0 g per day and 0.4 g per day, respectively). The proportion of
                                      67 This approach yields
tive risks as floating absolute risks (FARs).                         controls from developed countries who reported drinking no alco
floated standard errors (FSE) and floated confidence intervals        hol was 40%, and a furthar 28% reported consuming <5 g per
(FCI). Presentation of the resuits in this way enables valid compar   day, ie. lass than half a unit/drink of alcohol per day (Table 1).
isons between any two exposure groups, even if neither is the         Only about 1% of the controls from developed countrias reported
baseline group. Any comparison between groups must take the           drinking 35—44 g per day alcohol, i.a. about four units or drinks
variation in each estimate into account by summing the variances      daily, and a similar pro portion reported drinking       45 g per day.
of the logarithms of the two FARs.                                        Overall about half the women in developed countries reported
   To obtain comparability betwean the woman with breast cancer       that they had ever smoked, but smoking habits varied consider
and similar woman without breast cancar, all analyses ware routi      ably according to alcohol intake, both for cases and controls
nely stratified by study, and centra within study; by age (in single  (Table 1). Among controls from developed Countries who
years from 16 to 64, 65 to 69, 70 to 74, etc., up to 85 to 89); by    reported drinking no alcohol, 37% had ever smoked, and the
panty and, where appropriate, age whan the first child was bom        proportion of ever-smokers increased with increasing intake of
(nulliparous women ware assigned to a separate stratum and            alcohol, rising to 73% for controls who reported drinking
parous woman ware cross-classified according to panty (1—2, 3—         45 g per day alcohol (Table 1). The average alcohol consump
4, 5—6, 7+) and age at first birth (<20, 20—24, 25—29, 30+)).         tion reported by ever-smokers from developed countries was
Where appropriate analyses ralating to alcohol consumption ware       greater than that reported by never-smokers (8.4 g per day vs
stratified by smoking history (ever/never) and analyses relating to   5.0 g per day).
© 2002 Cancer Research UK                                                                 British Joumal of Cancer (2002) 87(11),  234— 1245
</pre>

====================================================================== Einde pagina 73 =================================================================

<br><br>====================================================================== Pagina 74 ======================================================================

<pre>                                                  Alcohol, tobacco and breast cancer
                          Collaborative Group on HomonaI Factors in Breast Cancer
1236
                Table 1 Reported alcohol and tobacco consumption among cases and controls in developed countries for whom information on both
                factors was available
                                                                                                    Alcohol consumption (g per day)
                                                            0                1—4              5—14          15—24       25—34         35—44          45+         Total
                CASES
                   Nu”,ber (%)                         18331 (36)        1 3785 (27)        10238 (20)    3444 (68)    2522 (5.0)    954 (1.9)    1 192 (2.4) 50466 (100)
                   Per cent that ever-smoked              39%                48%               58%           60%          56%          64%           70%          49%
                CONTROLS
                   Nu”ber (%)                          31872 (40)        22654 (28)         15484 (19)    5082 (6.3)   2727 (3.4)   II 19 (1.4)   1067 (1.3)  80005 (100)
                   Per cent that ever-sroked              37%                46%               55%           62%          60%          66%           73%          46%
     Table 2 Relative risk of breast cancer in relation to reported intake of                           amounts of statistical information, were grouped together as
     alcohol, according to smoking history                                                              ‘other’ in each of these categories. There was no strong evidence
                                                                                                        to suggest that the results varied substantially across studies
     g per day alcohol              Never-smoker             Ever-smoker           All women            52=60.7; P=0.3) or according to atudy design (y
                                                                                                        2
                                                                                                        (y                                                     2 for heteroge
     consumption                     relative risk           relative risk        relative risk
                                                                                                        neity=1.5; P=0.5). In the one study  52 which contributed data on
     (median)                              (FSE)                   (FSE)                 (FSE)
                                                                                                        alcohol, but not smoking, the estimated increase in the relative
     0 (0)                             1.00  (0.015)             .00 (0.018)         1.00  (0.012)      risk of breast cancer per additional 10 g per day intake was
     <5 (2)                            1.01  (0.020)           1.01  (0.020)         1.01  (0.014)       13.8% (SE 10.5%). Because of the large standard error, the esti
     5—14 (8)                          1.01  (0.023)           1.05  (0.021)         1.03  (0.015)      mated increase in relative risk in this study does not differ
     15—24 (18)                        1.19  (0.048)             .09 (0.035)         1.13  (0.028)      significantly from results for all other studies combined
     25—34 (29)                        1.22  (0.056)           1.19  (0.047)         1.21  (0.036)      »0.4, P=0.5).
                                                                                                        2
                                                                                                        (y
     35—44 (39)                        1.18  (0.093)           1.40  (0.077)         1.32  (0.059)
                                                                                           (0.060)
                                                                                                            The effect of adjusting for 11 other potential confounding
        45 (58)                        1 .49 (0. 1 10)         1 .46 (0.072)         1 .46
                                                                                                        factors (race, education, family history of breast cancer, age at
     lncrease in the relative                                                                           menarche, height, weight, body mass index, breastfeeding, use of
        risk of breast cancer
                                                                                                        hormonal preparations, and age at and type of menopause) on
        per 10 g per day (5E)         7.1% (1 .3%)            7.1% (0.9%)           7.1% (0.8%)
                                                                                                        the relationship in Figure 1 is shown in Table 3. Additional adjust
     CaIculated as floating absolute risk (FAR), with corresponding floated standard error              ment for each of these factors in turn did not materially alter the
     (FSE), and stratifled by study, age, panty, age St first birth and, for ‘all women’, by            magnitude of the increase in the relative risk of breast cancer asso
     smoking history (see Methods).                                                                     ciated with increasing levels of alcohol intake, suggesting that the
                                                                                                        associations in Figure 1 are not much confounded by any of them.
         Because alcohol and tobacco conaumption are so closely asso                                    Breast cancer in relation to tobacco consumption
     ciated, analyses of their effects were initially carried Out
     separately for never-smokers and ever-smokers (in the case of alco                                 Among the 22 255 cases and 40832 controls who reported drinking
     hol) and for drinkers and non-drinkers (in the case of tobacco).                                   no alcohol, the risk of breast cancer in ever-smokers did not differ
                                                                                                        significantly from that in never-smokers (relative risk for ever vs
                                                                                                        never-smokers=1.03, SE 0.023; NS). However, among women
     Breast cancer in relation to alcohol consumption
                                                                                                        who reported drinking alcohol, the findings for smoking were diffi
     Table 2 shows the relative risks and corresponding standard errors                                 cult to disentangle from the effects of the alcohol itself. When ever
     for breast cancer according to women’s reported daily intake of                                    smokers were compared to never-smokers the relative risk for
     alcohol for never-smokers and ever-smokers. In each group the                                      breast cancer was 1.09 (0,018) before stratification by the amount
     relative risk of breast cancer increased significantly with increasing                             of alcohol consumed, and declined to 1.05 (0.020) after stratifica
     intake of alcohol, increasing by the same amount, 7.1%, for each                                   tion. The corresponding y    2 value declined by three-quarters from
     additional 10 g per day intake of alcohol (P<O.00001 in each                                       23,4 to 6.4. Since alcohol consumption is known to be unreliably
     group). The trends associated with increasing levels of alcohol                                     68 and stratification for such a poorly measured variable
                                                                                                         measured,
     intake in never-smokers and ever-smokers did not differ signifi                                     reduced the 2 value by three-quarters, stratification by true alco
     cantly from each other (y              2 for heterogeneityO.OO2; P1.O).                            hol intake would be expected to reduce the z          2 value by even
     Therefore subsequent analyses concerning alcohol consumption                                        69 Since it is not possible to eliminate residual confounding
                                                                                                         more.
     include both never-smokers and ever-smokers, and the data are                                       among drinkers, results concerning tobacco consumption are
     stratified by smoking history as well as by study, age, panty and                                   restricted to women who reported drinking no alcohol at all, where
     age at first birth.                                                                                 such confounding should be minimised.
          When the data in smokers and non-smokers were combined the                                         The study-specific relative risks for breast cancer in ever-smokers
      relative risk of breast cancer increased with alcohol intake, increas                              compared to never-smokers are shown in Figure 3, for women who
      ing by 7.1% (SE 0.8%; P<0.0000I) for each additional 10 g per                                      reported drinking no alcohol. There was no marked variation in the
     day intake of alcohol, ie. for each extra unit/drink of alcohol                                     relative risk of breast cancer across studies (/252=58.0, P=0.3) or
     consumed on a daily basis (Figure 1). Compared to women who                                         study design (,2,=6.1, P=0.05). Information on current and past
     drank no alcohol the relative risk was 1.32 (0.059, P<0.00001)                                      smoking was available for all but five ’  2
                                                                                                                                                   studies,
                                                                                                                                                   2       3 (and two unpub
                                                                                                                                                           8
      for women whose reported alcohol consumption was 35—44 g                                           lished). Among ever-smokers in the remaining 48 studies 54% were
     per day and 1.46 (0.060, P’<0.00001) for a consumption of                                           current smokers and 46% were past smokers. Compared to oever
         45 g per day, where the average consumption was 57 g per day.                                   smokers the relative risk of breast cancer was 0.99 (SE 0.03) for
          The study-specific resuits are summarised in Figure 2, grouped                                 current smokers (Appendix III (www.bjcancer.com)), and 1.07
      according to study design. Studies which contributed the smallest                                  (SE 0.03) for past smokers (Appendix IV (www.bjcancer.com)).
      Bntith joumal of Cancer (2002) 87(11), 1234— 1245                                                                                               © 2002 Cancer Research UK
</pre>

====================================================================== Einde pagina 74 =================================================================

<br><br>====================================================================== Pagina 75 ======================================================================

<pre>                                                                                    Alcohol, tobacco and breast cancer
                                                                                    CollaborativeGrouponHornional Factors in Breast Cancer
                                                                                                                                                                   237
              1.7  —
              1.6
          ‘L. IZEIEZEEZZ.LZEZ.iEZZ.
                           0 (0)             10(1)                  20 (2)             30 (3)             40 (4)
                                                   Self-reported alcohol consumption, g per day (— number drinks daily)
                                                                                                                              50 (5)              60 (6)
Figure 1 Relative risk of breast cancer in relation to reported intake of alcohol. Relative risks are calculated as floating absolute risk (FAR) and stratified by
study, age, panty, age at first birth and smoking.
                                                                                                                                                                    1
   Among controls from developed countries a greater proportion                     calculated separately for various subgroups of women, subdi
of ever-smokers than never-smokers had had a bilateral oophor                       vided according to 15 personal characteristics including their
ectomy (8.7% s’s 7.6%) or a hysterectomy without bilateral                          age, childbearing pattern, race and familial patterns of breast
oophorectomy (13.3% vs 12.5%). The average age at bilateral                         cancer. Overall there was no significant variation in the trend
oophorectomy was 41.6 (SD 7.5) and 44.2 (SD 6.6), respectively                      associated with increasing intake of alcohol between categories
and the average age at hysterectomy was 38.6 (SD 9.3) and 40.0                      deflned by any of the 15 factors examined (Figure 4: global test
(SD 9.9), respectively. Average age at natural menopause was also                   for heterogeneity y2is=18.0; P=0.3). Nor was there significant
slightly earlier in ever-smokers than in never smokers, at 48.3 (SD                 variation in the relative risk of breast cancer associated with
4.8) and 49.3 (SD 4.7) years, respectively. The relative risk of                    having ever smoked across categories of the 15 characteristics
breast cancer in ever vs never-smokers was similar for women                        examined (Figure 4: global test for heterogeneity y2is=17.9;
who had had an oophorectomy, hysterectomy or natural meno                           P=0.3).
pause (Table 4) and additional stratification by age at and type                        Information on the extent of spread of the breast cancer was
of menopause did not materially alter the overall magnitude of                      available for about 60% of the study population. Both for
the relative risk (Table 3). Nor did additional stratification by                   tumours localised to the breast and for tumours that had
10 other potential confounding factors much alter the relative                      spread beyond the breast, the risk of breast cancer increased
risk.                                                                               with increasing alcohol consumption (increase in relative risk
            363 that together inciuded a total of 4781 cases
    Eleven studies’                                                                 of breast cancer of 6.9% (1.3%) and 9.4% (1.5%), respectively,
and 12713 controls, contributed data to this collaboration on                       per 10 g per day alcohol consumption: y2=33; P=0.07). There
tobacco consumption for each woman, bot not on alcohol                              was no signiflcant difference in the extent of tumour spread
consumption. The combined relative risk of breast cancer in                         among the cases according to tobacco consumption (721=3.0,
ever-smokers compared to never-smokers in these 11 studies was                      P=0.08).
1.05 (SE 0.05), bot because of the potential for confounding by
alcohol the results from these studies have not been inciuded in                    Cumulative incidence of breast cancer
the main analyses.
                                                                                    Around 1990 the cumulative incidence of breast cancer up to age
                                                                                    80 years was between about eight and 10 per 100 women in devel
Consistency of the fi.ndings
                                                                                             ° The average consumption of alcohol by
                                                                                    oped countries.
                                                                                             64657
The increase in the relative risk of breast cancer for each addi                    controls studied here from developed countries was 6.0 g per
tional 10 g per day intake of alcohol consumption was                               day. 1f the dose-response relationship described here is valid, it is
© 2002 Cancer Research UK                                                                                British Joumal of Cancer (2002) 87(ll), 1234— 1245
</pre>

====================================================================== Einde pagina 75 =================================================================

<br><br>====================================================================== Pagina 76 ======================================================================

<pre>                                          Alcohol, tobacco and breast cancer
                        Collaborative Group on Hormonal Factors in Breast Cancer
1238
                                                      Number        Mean Intake of        % lncrease In             % lncrease In relative risk of
                                                                    alcohol (glday) relative risk per 1 Og!day       breast cancer per 1 Og!day
      Study(Country)rel.                        Cases/Controls      Cases!Controls lntake of alcohol (SE)            Intake of alcohol & 99% Cl
      COHORT STUDIES:
      Nurses Health Study(USA)
                       8
      Cariadian NBSS(Canada)
                   24
                                                   2870/11480
                                                     753/2857
                                                                         6.3/5.2
                                                                         9.2/8.9
                                                                                             4.4 (2.5)
                                                                                              1.2 (4.1)
                                                                                                                                  1-
                                                                                                                                  H—
      American Cancer Society(USA)
                          51                        1196/4829            9.5/9.0             7.0 (5.3)
      Netheriands Cohort(NetherIands)                470/1686            6.3/5.8             2.8 (6.9)
      lowa Womens I-fealth(USA)
                       19                           1188/4752            4.1/3.5              8.1 (4.8)
      Million Women Study(UK)
                       47                           1436/5744            5.6/5.2              9.8 (6.3)
      Otherliawat                                   1780/7083            0.8/0.4            34.7 (41.5)
     All cohort studies                            9693/38431            5.0! 4.4            5.0(1.7)
      CASE-CONTROL, POPULATION CONTROLS:
      3
      Brinton(USA)                                  1726/2179            8.7/7.2              3.7 (4.0)                           .4—
      °
      1
      CASH(USA)                                     4455/4672            8.2/7.4             3.5 (2.8)
      Bernstein(USA)                                  676/676            6.9/5.6             10.0(7.4)
      15
      BaIn/Sisklnd(Australla)                         487/981            6.7/5.6             -1.2 (9,7)
      8
      Rohan(Australia)                                451/451            7,0/4.5            4.9 (10.6)
      17
      Ewerlz(Denmark)                               1525/1 398           7.4/7.2             4.3 (5.0)                      —
      Long lsIand(USA)                              1183/1184            5.9/4.6             15.8(7.7)
      16
      Clarke(Canada)                                 607/1214            8.5/7.6              3.6 (6.8)                  —
      PauI/Skegg(New Zealand)
                         18                          888/1857            3.8/3.4            7.3 (11.4)
      DaIing(USA)                                     747/961            7.4/6.8             -0.6 (5.2)
      31
      Ross/Paganlnl-HiII(USA)                       1055/1092            7.7/6.5              8.6 (5.3)
                           npbiis5od
      UK studies(UK)
           282                                      1871/1871             5.6/5.8             1.2(5.1)
     4 State Study(USA)
               °
               3                                    6880/9525             6.6/5.3            13.4 (2.2)
      Rookus/van Leeuwen(Netherlands)
                     27                               918/918            10.6/9.6             5.6 (4.8)
      22
      Yang/Gallagher(Canada)                        1019/1025             4.9/5.2             1.1 (9.0)
      PrimicJZakeIj(SIovenia)                         619/619            3.7/2.0            14.5 (13.8)                           —.
      Stanford/HabeI(USA)                             450/492             6.4/5.4            -1.0(9.8)                            —.
     °
     4
     WISH(USA)                                      1866/2007             6.0/6.6            -1.8(5.4)                         1  --
      46
      Magnusson(Sweden)                             3168/3285             2.7/2.2            19.7(8.4)
      45
      ’
      42
      McCredie/i-Iopper(Australia)                  1581/1021             6.9/6.0             8.3 (6.4)
      °
      4
      Chang-Claude(Germany)                          656/1283            10.8/8.5            14.8 (5.0)
     49
     Johnson(Canada)                                2338/2427             5.2/6.2             6.6 (4.5)
      9
      ,
      7
      5
      Other
      2
      4
      ’
      37
      ’’          0
                  1
                  5                                 3509/4656             2.1/2.4             9.4 (7.0)                             1.
      All case-control, pop controls
      CASE-CONTROL, HOSPITAL CONTROLS:
                                                  38675145794            6.0/5.3              7.4(1.1)
                                                                                                                                   4>
     12
     Vessey(UK)                                     1125/1125             3.1/3.5            -7.4 (7.6)
      Franceschi(italy)
      443                                           2929/2963           14.2/12.2             3.2 (2.6)                           Ii-
      L/Gerber/CIavei(France)2l439                  1204/1724             8.9/6.4            20.6 (6.3)
      La Vecchia(italy)
          ’
          1                                         3623/2729           17.3/14.1             9.5 (2.5)
      25
      Katsouyanni(Greece)                            795/1548             4.4/4.1           20.0(13.7)
      ’
      23
      Other                                           471/753            14.5/8.5             9.7 (8.4)
      All case-controi, hospital controis         10147/10842           12.5/9.4              7.3(1.7)
                ALL STUDIES                      58515/95067             7.0/5.4              7.1 (0.8)
                                                                                                                -25%             0%          25%         50%
     Figure 2 Details of and results from studies on the relation between alcohol consumption and breast cancer. Relative risks are stratified by age, par-ity, age
     at first birth and smoking history.
     Bntish Joumal of Cancer (2002) 87(11), 1234— 1245                                                                           © 2002 Cancer Research UK
</pre>

====================================================================== Einde pagina 76 =================================================================

<br><br>====================================================================== Pagina 77 ======================================================================

<pre>                                                                          Alcohol, tobacco and breast cancer
                                                                          Collaborative Group en Horrnonal Factors in Breast Cancer
                                                                                                                                                   1239
                                                Number         % ever       Relative risk of breast          Relative risk of breast cancer
                                                              snioked       cancer In over- versus          In over- versus never-smokers
Study(Country)t                            CaseslControls  Cases!Controls     never-smokers (SE)                        & 99% CI
COHORT STUDIES:
Nurses Heaith Study(USA)
                  8                           1224/5599         49/49              1.01(0.07)                            —II--
Canadian NBSS(Canada)
            24                                   181/662        35/35              1.25 (0.23)                        —
American Cancer Society(USA)
                    51                          213/922         34/33              1.07 (0.19)
Netherlands Cohort(NetherIands)                  119/504        27/30              0.89 (0.23)
lowa Womens Health(USA)
                  19                           679/2765         25/26              0.93 (0.10)                        -u
Million Women Stucly(IJK)
                  47                           324/1291         50/44              1.24(0.15)
        13,,41
1
Other                                          1923/7655         4/5               0.78 (0.12)
All cohort studies                           4663119398         25/26              1.00 (0.04)                            < >
CASE-CONTROL, POPULATION CONTROLS:
3
8rinton(USA)                                    649/872         29/26              1.12(0.14)
 °
 1
 CASH(USA)                                     181 7/1 821      49/43              1.28(0.08)
 44
 Bemstein(USA)                                  336/317         50/48              1.18(0.20)                         —
 15
 Bain/Siskind(Austraiia)                        248/514         32/29              1.31 (0.26)                        —
 8
 Rohan(Australia)                                188/213        35/32              1.06(0.31)
 17
 Ewertz(Denmark)                                227/198         59/57              0.88 (0.27)
Long tsland(USA)
       38                                       153/208         37/34              0.99 (0.32)
16
Clarke(Canada)                                   114/211        40/42              0.88(0.31)
 PauI/Skegg(New Zealand)
                    18                         538/1058         43/41              1.09(0.13)                           —.
 29
 Daling(USA)                                    211/286         42/42              0.87(0.21)                            —
 31
 RossfPaganlni-HllI(USA)                        578/590         53/52              1.02 (0.13)
 UK 2studies(UK) unpubiish                      655/662         47/45              1.08 (0.13)                          —.
4 State Study(USA)
         °
         3                                     1507/2247        39)39              1.07 (0.09)                           -1-
 Rookus/van Leeuwen(Netherlands)
               27                               247/247         52/51              0.90 (0.21)                            .
YangIGaIlagher(Canada)                          505/517         48/44              1.15(0.17)                          —
 32
 Primic/Zakelj(Slovenia)                         115/128        29/30              0.67 (0.38)
Stanford/HabeI(USA)                              152/181        52/49              0.79 (0.26)
°
4
WISH(USA)                                       353/241         59/68              0.63(0.21)
46
Magnusson(Sweden)                              131 1/1 312      32/33              0.91 (0.08)
4
’
42                5
McCredie/Hopper(Australia)                      774/518         38/36              1.03(0.15)
 °
 4
 Chang-Claude(Germany)                           168/251        46/52              0.94 (0.25)
 49
 Johnson(Canada)                                974/1110        42/40              1.14 (0.11)
 5792021
 Other                                        2851/3567         11)13              0.99(0.12)                         -
 All case-control, pop controls              14671/17269        36/35              1.07(0.03)
 CASE-CONTROL, HOSPITAL CONTROLS:
 12
 Vessey(UK)                                      154/171        44/53              0.71 (0.30)
 Franceschi(Itaiy)
 ’
 4                                              831/1025        31/31               1.01 (0.12)
 L/Gerber/Ciavei(France)2l4                      492/923        18/24              0.82 (0.16)
 La Vecchia(ItaIy)                              980/1034        28/30              0.82 (0.10)
 25
 Katsouyanni(Greece)                             219/462        21/24               1.28 (0.29)                      —
 23
 Other                                           245/550        20/26              0.72 (0.25)
 All case-control, hospital controls          2921/4165         27/29              0.89 (0.06)
         ALL STUDIES                        22255140832         33/30              1.03(0.02)                                 >
                                                                                                      0         0.5         1.0       1.5      2.0
Figure 3 Details of and results on the relation between tobacco consumption and breast cancer in woman who reported dnnking no alcohol. Relative
riss are stratified by age, pahty and age at first birth.
© 2002 Cancer Research UK                                                                      British Joumal of Cancer (2002) 87(11), 1234— 1245
</pre>

====================================================================== Einde pagina 77 =================================================================

<br><br>====================================================================== Pagina 78 ======================================================================

<pre>                                              Alcohol, tobacco and breast cancer
                         Collaborative Group on Hormonal Factors in Breast Cancer
l240
            Table 3 Effect of additional adjustrnent for vahous factors on the relative nsk of breast cancer associated with alcohol and
            tobacco consurnption
                                                                                                                         Relative risk (SE) of breast cancer in
                                                                Per cent increase (SE) in the relative risk of       ever-smokers, compared to never-smokers
                                                               breast cancer per 10 g per day alcohol intake for women who reported drinking no alcohol
            After stratification for study, age, panty, age at                   7. 1% (0.8%)                                            1.03 (0.02)
              first birth and, for analyses concerning
              alcohol, tobacco consurnption
            After additional stratification fon
              race                                                               7.2%  (0.8%)                                               .03 (0.02)
              education                                                          7.3%  (0.8%)                                            1.04   (0.03)
              mother or beter witt breast cancer                                 7.2%  (0.8%)                                               .02 (0.03)
              age at menarche                                                    7.4%  (0.8%)                                            1 .04  (0.03)
              height                                                             7.5%  (0.8%)                                            1 .02  (0.03)
              weight                                                             7.2%  (0.8%)                                             1 .04 (0.03)
              body mass index                                                    6.9%  (0.8%)                                             1 .04 (0.03)
              breastfeeding                                                      6.9%  (0.8%)                                             1.02  (0.02)
              ever use of hormonal contraceptives                                6.6%  (0.8%)                                             1 .02 (0.03)
              ever use of hormone replacement therapy                            7.3%  (0.8%)                                             1 .02 (0.03)
              type of and age at menopause                                       7.2%  (0.8%)                                             1 .06 (0.03)
     estimated that about 4% of breast cancers in developed countries                         the effects of other factors. Potential confounding by age, study,
     are attributable to alcohol. The cumulative incidence of breast                          panty, age at first birth and tobacco consumption were minimised
     cancer by age 80 years is estimated to increase from 8.8 per 100                         by stratification. Ever-smokers had their natural menopause about
     women in non-drinkers to 9.4, 10.1, 10.8, 11.6, 12.4 and 13.3,                           1 year earlier, on average than never-smokers and were also more
     respectively, per 100 women consuming an average of 1, 2, 3, 4,                          likely to have had a bilateral oophorectomy or hysterectomy, but
     5 and 6 alcoholic drinks each day (see Figure 5). In developing                          adjustment for type of and age at menopause had little effect on
     countries, where alcohol consumption is very low, averaging only                         the relative risk of breast cancer in ever- vs never-smokers (Tables
     about 0.4 g per day, alcohol would make a negligible contribution                        3 and 4). In addition, possible confounding by race, education,
     to the total number of cases of breast cancer.                                           family history of breast cancer, age at menarche, height, weight,
                                                                                              body mass index, breastfeeding and use of hormonal preparations
                                                                                              was examined by adjustment for each factor in turn, but none
     DISCUSSION
                                                                                              materially altered the estimates of relative risk (Table 3). Since
     There is potential for confounding between the possible effects of                       the relative risk estimates for breast cancer in relation to both alco
     alcohol and of tobacco on breast cancer, as drinking and smoking                         hol and tobacco consumption did not appear to differ substantially
     are closely associated, one with another. Among controls from                            according to any of these factors, there is no strong evidence for
     developed countries, the proportion of ever-smokers rose from                            interaction between either of these exposures and the 15 factors
     37% in women who reported drinking no alcohol at all, to 73%                             examined (Figure 4).
     in women drinking ?‘ 45 g per day alcohol, and alcohol consump                               There was no significant difference in the extent of tumour
     tion was greater in ever-smokers than in never-smokers, averaging                        spread according to either alcohol or tobacco consumption,
     8.4 and 5.0 g per day, respectively.                                                     suggesting that there is little differential detection of breast cancer
        The relative risk of breast cancer was found to increase with                         or effect on tumour growth by these exposures.
     increasing intake of alcohol, both in never-smokers and in ever
     smokers, and the magnitude of the increase was the same in each                          Combining resuits from different studies
     group (an increase of 7.1% in the relative risk of breast cancer for
     each additional 10 g per day alcohol; 95% CI 5.5—8.7%                                    Combining results across many studies bas the advantage of yield
     P<0.00001 overall). The observed association between breast                              ing estimates of the relative risk that are not subject to as much
     cancer and alcohol consumption is therefore unlikely to be an
     indirect effect of tobacco.
        Conversely, the relationship between smoking and breast
                                                                                              Table 4 Relative nsk of breast cancer in ever vs never smokers.
     cancer was found to be confounded by alcohol. Among women
                                                                                              according to menopausal status, in women who reported drinking no
     who drank no alcohol, ever-smokers and current smokers were                              alcohol. Relative nsks are stratified by study, age. panty and age at first birth
     not at an increased risk of breast cancer compared to never
     smokers. Among women who drank alcohol, however, adjust                                  Menopausal status                                            Relative risk (SE)
     ment of the relative risk of breast cancer by the amount of
     alcohol consumed had a substantial effect on the resuks and,                             Pre”enopausal                                                    1.07 (0.05)
     since it is not possible to measure alcohol intake reliably and                          Natural rrenopause
     thus eliminate residual confounding due to alcohol, we chose                                before age 45 yeans                                           1.1 1 (0.15)
                                                                                                 at age 45—49 years                                            0.98 (0.08)
     to base our assessment of the effect of tobacco 00 breast cancer
                                                                                                 at age 50 years                                               1.12 (0.06)
     on the 22 255 cases and 40 832 controls recorded as drinking no                          Bilateral oophorectomy
     alcohol at all. In this large group of women the resuits suggest                            before age 45 years                                           0.78 (0.1 6)
     that smoking has littie or no independent effect on the risk of                             at age 45 years                                               0.82 (0.15)
     developing breast cancer.                                                                Hysterectomy before menopause                                    1 .08 (0.09)
        The association between breast cancer and alcohol or tobacco
     consumption does not appear to be materially confounded by                               726  for heterogeneity7.5; P0.9
     Bntish joumal of Cancer (2002) 87(11),              234— 1245                                                                                © 2002 Cancer Research UK
</pre>

====================================================================== Einde pagina 78 =================================================================

<br><br>====================================================================== Pagina 79 ======================================================================

<pre>@
    50
o
r-)  15
n
                                                                                   ALCOHOL                                                                            TOBACCO
                                                           % lncrease in relative risk                                                            Relative rlsk of breast cancer
(0                                                         per lOg of alcohol per day                                                                  In ever smokers versus
03   01
                                                                     & 99% Cl                               % increase (SE)                          never smokers & 99% Cl              RR (SE)
D
         ALL WOMEN                                                                                          7.1 (0.8)                                                 1-                 1.03 (0.02)
c
         AGE AT DIAGNOSIS
           <50                                                                                                                                                        1—                 1.05 (0.04)
                                                                                                            7.7(1.0                                              .    —                  1.01 (0.03)
     —     50                                                                                               6.2(1.9
     n
     53
         PARITY
                                                                                                            7.311.9                                                   —                  0.99(0.061
     n     Nutiparous
     !i(   Parous                                                                                           7.0 (0.9                                                  1-                 1.03(0.03)
     5   AGE AF FiRST BIRTH
           <25                                                                                                                                                        1—                 1.05(0.03)
                                                                                  •.                        5.9 (1.3                                                  —                  1.02 (0.04)
           25                                                                                                   .I
                                                                                                                1
                                                                                                                (
     o   BREASTFEEEtING
     D     Ever                                                                 •                           5.0 1.6                                                                      0.99 ‘0.041
           Never                                                                                            85 16                                                                        1.08(0.05)
     2   RACE
           while                                                                —,--—
                                                                                                            7.5 (l.0                                                                     1.07 (0.031
     0                                                      --                                                                                                        —                  0.95 (0.08)
           other                                                                                             2.2 (5. )
         COUNTRY
           developed                                                                                        7.1 (0.8)                                                                    1.03 (0.02)
     o     developing                                       --                                        --     2.3(13.5)                                                !:_—               0.86 0.11)
         EDUCA11ON
           <13 years                                                                                        8.3(1.9                                                                      1.09 (0.041
           l3years                                                                                          5.8(1.3                                                                      1.00(0.045
     n
     °
         MOTHER OR SISTER WfllI BREAST CANCER
           yes                                                                                        --    12.2(4.1)                                                                    1.20(0.121
     8     no                                                                   —1——       •                6.9 (l.9)                                                                    lol (0.03)
         AGEATMENARCHE                                                                                                                                                                                       0
                                                                                                                                                                         —
           <13                                                                                              9.0(1                                                                        1.09(0.041      o
    -g
                                                                              —--—                                                                                    —
     —.    13                                                                                               6.4(1.1.4                                                                    1.02 (0.035
     o
         HEIGHT                                                                     •
                                                                                                                                                                                                         <   0
           <165 cm                                                                                          8.9(1.9                                                   —                  1.01 (0.03’
     S
     <     1 65cm                                                                                           5.8(1.3                                                   1—                 1.05(0.04)
                                                                                                                                                                                                         -,
                                                                                                                                                                                                             ri
                                                                                    •                                                                                                                    °   n
         WEIGHT                                                                                                                                                                                              0
     o     <65kg                                                                                            7.6 (1.1                                                                     1.06(0.041
           65 kg                                                                                            6.7 (i .5                                                 1—                 1.04 (0.045     o
         BODY MAS INDEX
                                                                               ———                          6.8(1.0)                                                                     1.06 (0.03)     o   .
    1
    a      <25 kg/rn
           25 kg/m
               2                                                                                            7.2 (1.7)                                                 1—                 1.04 (0.04)         15
         EVER USE OF HORMONAL CONTRACEPTIVES
           yes                                                                      •                       5.6 (1.9                                                  1—                 1.05(0.041
           no                                                                                               7.2(1.1                                                   —                  1.01(0.03)      -fl 53
         EVER USE OF HORMONE REPLACEMENT ‘FHERAPY
     0                                                                                                      6.6 (2.l                                                                     1.09(0.061
           yes                                                             —s——————
     <     no                                                                                               6.5 (1.0                                                  1—                 1.03(0.03)      —.
D
         MENOPAUSALSTATIJS
           premenopausal                                                                                    6.3 (1.41                                                                    1.07 (0.05)
           postmenopausal                                                       •                           8.1 (1.35                                                                    1.01 (0.04)
0
0                                                                                                                                                                                                        —
                                                                                                                                                                                                         n
                                                                                                                                                                                                         5,
                                                         -10%           0%            10%             20%                                          0.5             1.0        1.5                        S
                                                                                                                                                                                                         0
                                                                                                                                                                                                         (0
—                                             *                                                                                      *
                                                stralitied by study, age, panty, age at first birth and tobacco consumption.           stratified by study, age, panty and age at first birtti; analyses
—                                                                                                                                      restricted to women who reported dnnking no alcohol.
r.J
                                                                                                                           Epidemiology
</pre>

====================================================================== Einde pagina 79 =================================================================

<br><br>====================================================================== Pagina 80 ======================================================================

<pre>                                         Alcohol, tobacco and breast cancer
                       Collaborative Group on Hormonal Factors in Breast Cancer
1242
         0
         e  14                                                                  risk breast cancer (refereoce oumber 81, in Appeodix II
         S                                                  13.3 6 drinks/day
         0                                                                      (www.bjcaoeer.com)). Nooe of these studies has, however,
         S
        0   12                                                                  published results on the risk of breast eaocer in relatioo to smok
        0                                                        4 drinke/day
                                                                                ing, restricted to womeo who oever drank alcohol.
         S
         0.
            10                                                   2 drinks/dey
         5,
         u
         5,                                                                     Limitations of these findings
         5,                                                       no alcohol
         3
         5
             8                                                                  Overall, the relative risk of breast cancer appeared to iocrease by
         5
        .0
                                                                                7.1% (95% CI 5.5—8.7%) for eaeh additiooal 10 g per day iotake
         0
             6                                                                  of alcohol i.e. for each extra unit/driok of alcohol coosumed 00 a
         5)
         3                                                                      daily basis. Information oo alcohol eoosumption was, however,
         5,
         5)
             4                                                                  usually self-reported, deseribiog driokiog habits at around the time
         .3
         5,
                                                                                that the womeo were ioterviewed. No information oo the pattero
         8                                                                      of intake, iocludiog the type of alcohol coosumed and the duratioo
         5,                                                                     of intake, was collected for this collaboratioo. There is no stroog
         5)
         S                                                                      evideoce here to suggest biased reportiog of alcohol eoosumptioo
         5)
        t)                                                                      in case-cootrol studies, sioce there was no sigoificant differeoce
               20 25 30 35 40 45 50 55 60 6        70 75 80
                                                                                in results between case—cootrol and cohort studies (iocreases of
                                     Age
                                                                                7.4% and 5.0% per 10 g per day, respectively; y’ for heterogene
     Figure 5 Estimated cumulative incidence of breast cancer per 00
                                                                                ity=1.5, P=0.2). However, self-reported ioformatioo oo alcohol
     women in developed counthes, according to the number of alcoholic          eonsumptioo is koowo to uoderestimate true coosumption.
                                                                                                                                      65
     drinkt consumed each day (see Methods).                                    Systematic under-reportiog of consumptioo by both cases and
                                                                                cootrols would result io ao overestimation of the relative risk of
                                                                                breast cancer for a giveo level of alcohol eoosumptioo. By eootrast,
     random fluctuation as that found in any individual study. The              random miselassification amoog both cases and cootrols would
     studies that contributed to these findiogs were of different designs       have the opposite effect, resultiog in ao underestimation of the
     and inciuded women with a wide range of alcohol and tobacco                relative risk. These two types of measurement error are inevitable,
     consumption and of other personal characteristics. Nevertheless,           but counter-actiog, and it is oot possible to estimate their overall
     the relatiooships between breast cancer and alcohol and tobacco            effect on the relative risks calculated here. Moreover, the shape
     were seen consistently across studies and study designs, and for           of the dose-respoose relatiooship could be changed if, for example,
     women of different ages, different childbearing histories, and for         heavy drinkers were more likely to uoder-report intake than
     women who differed according to varinus other personal eharacter           moderate drinkers. Taken together, these reportiog errors imply
     istics. The results were not unduly influenced by any particular           that some uncertainty remaios about the true quaotitative effect
     study or groups of studies.                                                of an intake of a fixed amouot of alcohol oo the risk of developiog
          Because of the strong association between alcohol and tobacco         breast eaocer.
     consumption, the main analyses were restricted to data from the                The true relatiooship betweeo alcohol coosumptioo aod breast
     53 studies in which information on both exposures had been                 cancer might, perhaps, be more eurved thao is suggested by the
     collected in the same women. Results from the only study      52 that      shape of the relationship showo in Figure 1, beeause of miselassi
     had provided individual data on alcohol, but not tobacco, did              fication of alcohol intake, as may also have occurred with cigarette
     not differ significantly from the overall findings for breast caoeer       smoking and lung caocer.
                                                                                                     71 Any firm cooclusioo about the risk of
     and alcohol. The remaining 11 studies5363 that provided indivi             breast caocer at low levels of alcohol intake is, however, prohibited
     dual data on tohaeco, but not oo alcohol, could not contribute             by the likelihood of measuremeot errors, particularly the teodeocy
     directly to this review, since it was not possible to take into            for uoderestimatioo of the amouot druok, aod by the possibility
     account for the important eonfounding effect of alcohol. None              that non-drinkers may differ in some relevant, but uomeasured,
     of the publications from these 11 studies has, however, claimed            ways from those who sometimes drink alcohol. Heoce, the possibi
     that smoking affected the risk of breast eaocer.                           lity of a threshold dose of alcohol canoot be reliably assessed from
         As far as cao be ascertained, over 80% of the worldwide epide          the data io Figure 1.
     miological data that have been assembled no the relatiooship                   These results provide oo direct evideoce about possible mechao
     between breast cancer and alcohol and tobacco coosumption were             isms of carcioogeoesis by alcohol on the breast. There is, however,
     contributed to this eollaboration. Aoother 20 studies were identi          aecumulatiog evideoee that regular iotakes of moderate amounts of
     lied with relevant data that together iocluded about 12000                 alcohol affect sex hormooe levels. For example, the results of a
     women with breast cancer (see Appeodix II (www.bjcancer.com)),             receotly published small randomised trial of 51 postmeoopausal
     but beeause results were presented in a differeot way in eaeh study,       women suggested that sex hormooe levels may be inereased after
     it is difficult to combine the published data direetly. Nevertheless,      the eoosumptioo of 30 g per day alcohol for 8 weeks,72 levels of
     out of the six largest studies all but one (refereoce oumber 66, in        coosumptioo that are assoeiated here with a dear excess risk of
     Appeodix II (www.bjcaoeer.com)) reported a statistically sigoificant       breast eaocer.
     inereased risk of breast eaocer with iocreasiog iotake of alcohol.             With respect to the coosumptioo of tobaeeo, the main exposure
     Eaeh of these six studies ineluded at least 500 womeo with breast          variable examioed here was whether or not a womao had ever
     caneer and altogether they eomprised most of the information that          smoked. No information was colleeted for this eollaboratioo 00
     had not been cootributed io this collaboratioo. The remaioiog 14           the amouot smoked or on the age that smoking started or stopped,
     studies were eomparatively small and oooe of their published               oor bas attention been given to the reported effeets of eoviroomeo
     results oo alcohol differed substaotially from those reported here.                            49 as active smoking ooly bas been
                                                                                tal exposure to tobaeco,
                                                                                                    ’
                                                                                                    35
     Therefore the fiodiogs on alcohol and breast eaoeer from studies           eoosidered. Although some past smokers may have smoked rela
     not ineluded here do not appear to differ materially from these            tively iofrequeotly, eurreot smokers are likely to have had
     results.                                                                   substaotial lifetime exposures to tobaeeo, particularly in e000tries
          Ooly one of the 20 studies that had not eootributed to this           where lung eaocer rates in women are high. Just over half the
     collaboration claimed that smoking is associated with ao iocreased         ever-smokers ineluded in these analyses were eurreot smokers,
     Bntish joumal of Cancer (2002) 87(11),      234— 1245                                                                © 2002   Cancer Research UK
</pre>

====================================================================== Einde pagina 80 =================================================================

<br><br>====================================================================== Pagina 81 ======================================================================

<pre>                                                                               Alcohol, tobacco and breast cancer
                                                                               Collaborative Group on Honnonal Factors in Breast Cancer
                                                                                                                                                              1243
and among them the risk of breast cancer was similar to that in                cancer is estimated to increase by about 0.7 per 100 women for
never-smokers (relative risk=0.99 (95% CI, 0.96—1.03)). The find               each extra unit or drink of alcohol consumed on a daily basis.
ings from case—control studies could, in theory, be biased if                  For example, the cumulative incidence of breast cancer by age 80
women with breast cancer stopped smoking when they first devel                 years is estimated to increase from 8.8 per 100 women who drink
oped symptoms, or if there were differential reporting of smoking              no alcohol to 10.1 or 100 who consume two alcoholic drinks daily
by cases and controls. However, the resuits from cohort studies,               and to 11.6 per 100 who consume four drinks daily. This excess
where exposure information was collected prospectively, suggest                risk should be considered in the context of the beneficial effects
no increase in the risk of breast cancer in ever-smokers or current            of alcohol, in moderation, on cardiovascular disease, and its harm
smokers compared to never-smokers (relative risk=1.00, 95% Cl                  ful effects on cirrhosis and on cancers of the mouth, larynx,
0.93— 1.07, for ever-smokers; and =0.94, 95% CI 0.84— 1.05, for                oesophagus and the liver.74
                                                                                                        ’
                                                                                                        75
current smokers).
Public health implications
                                                                               ACKNOWLEDGEMENTS
1f the pattern of breast cancer associated with increasing levels of
alcohol consumption estimated here is valid, then about 4% of                  This review would not have been possible without the tens of thou
the breast cancers in women in developed countries may be attri                sands of women with and witbout breast cancer who took part in
butable to alcohol. The consumption of alcohol by most women in                this research. Central pooling, checking and analysis of data was
developed countries is relatively low, with about two-thirds                   supported by the Cancer Research UK and the UNDP/UNFPA/
consuming little or no alcohol each day. For women in developed                WHO/World Bank Special Programme of Research, Development
                                                                                                                                                                1’
countries who regularly drink alcohol, the lifetime risk of breast             and Research Training in Human Reproduction.
REFERENCES
  1 Hiatt RA, Bawol RD (1984) Alcoholic beverage consumption and breast        18 Sneyd MJ, Paul C, Spears GF, Skegg DCG (1991) Alcohol consumption
    cancer incidence. Atn 1 Epideniiol 120: 676—683                                and risk of breast cancer. Int 1 Cancer 48: 812—815
  2 Lê MG, Moulton LH, Hili C, Kramar A (1986) Consumption of dairy            19 Gapstur SM, Potter JD, Sellers TA, Folsom AR (1992) lncreased risk of
    produce and alcohol in a case-control study of breast cancer. 1 Nat!           breast cancer with alcohol consumption in postmenopausal women.
    Cancer lost 77: 633—636                                                        Ant 1 Epidentiol 136: 1221—1231
  3 Harvey EB, Schairer C, Brinton LA, Hoover RN, Fraumeni Ir 1F (1987)        20 Ursin G, Aragaki CC, Paganini-Hill A, Sierniatycld j, Thompson WD,
    Alcohol consumption and breast cancer. 1 Nat! Cancer lost 78: 657—661          Haile RW (1992) Oral contraceptives and premenopausal bilateral breast
  4 La Vecchia C, Decarli A, Parazzini F, Gentil A, Negri E, Cecchetti G, Fran     cancer: a case-control study. Epidentiology 2: 414—419
    ceschi S (1987) General epidemiology of breast cancer in Northern Italy.   21 Wang Q-S, Ross RK, Yu MC, Ningj-P, Henderson BE, Kimm HT (1992)
    Int 1 Epideiniol 16: 347—355                                                   A cate-control study of breast cancer in Tianjin, China. Carver Epidentiol
  5 Lee NC, Rosero-Bixby L, Oberle MW, Grimaldo C, Whatlev AS, Rovira              Biontarkers Prevent 1: 435—439
    EZ (1987) A case-control study of breast cancer and hormonal contracep     22 Yang CP, Daling JR, Band PR, Gallagher RP, White E, Weiss NS (1992)
    tion in Costa Rica. 1 Nat! Cancer lust 79: 1247—1254                           Non contraceptive hormone use and risk of breast cancer. Cancer Catises
  6 Willen WC, Stampfer Ml, Colditz GA, Rosner BA, Hennekens CH, Spei              Control 3: 475—479
    zer FE (1987) Moderate alcohol consumption and the risk of breast          23 Ferraroni M, Gerber M, Decarli A, Richardson S, Marubini E, Crastes de
    cancer. N Eng! 1 Pvled 316: 1174— 1180                                         Paulet A, Pujol H (1993) HOL-cholesterol and breast cancer: a joint study
  7 Adami HO, Lund E, Bergstrom R, Meirik 0 (1988) Cigarette smoking,              in Northern ltaly and Southern France. Int 1 Epidemiol 22: 772—780
    alcohol consumption and risk of breast cancer in young women. Br 1         24 Friedenreich CM, Howe GR, Miller AB (1993) A cohort study of alcohol
    Cancer 58: 832—837                                                             consumplion and risk of breast cancer. A,n / Epideiniol 137: 512—520
  8 Rohan TE, McMichaelAj (1988) Alcohol consumplion and risk of breast        25 Katsouyanni K, Trichopoulou A, Stuver S, Vassilaros S, Papadiamantis Y,
    cancer. Int / Cancer 41: 695—699                                               Bournas N, Skarpou N, Mueller N, Trichopoulos D (1994) Ethanol and
  9 Yuan J-M, Yu MC, Ross RK, Gao Y-T, Henderson BE (1988) Risk factors            breast cancer: an association that may be both confounded and causal.
    for breast cancer in Chinese women in Shanghai. Cancer Res 48: 1949—           Int J Cancer 58: 356—361
    1953                                                                       26 Land CE, Hayakawa N, Machado SG, Yamada Y, Pike MC, Akiba S,
10 Chu SY, Lee NC, Wingo PA, Webster LA (1989) Alcohol consumption                 Tokunaga M (1994) It case-control interview study of breast cancer
    and the risk of breast cancer. Ani 1 Epidenziol 130: 867—877                   arnong Japanese A-bomb survivors. 11. Interactions with radiation dose.
11 La Vecchia C, Negri E, Parazzini F, Boyle P, Fasoli M, Gentile A, Fran          Cancer Catises Control 5: 167—176
    ceschi S (1989) Alcohol and breast cancer: update from an Italian case     27 Rookus MA, van Leeuwen FEfor the Netherlands Oral Contraceptives and
    control study. Eur 1 Cancer Clin Oncol 25: 1711— 1717                          Breast Cancer Study Group (1994) Oral contraceptives and risk of breast
12 Meara J, McPherson K, Roberts M, Jones L, Vessey M (1989) Alcohol,              cancer in women aged 20—54 years. Lancet 344: 844—851
    cigarette smoking and breast cancer. Br 1 Cancer 60: 70—73                 28  Smith SJ, Deacon JM, Chilvers CE (1994) Alcohol, smoking, passive
13 MilIs PK, Beeson WL, Phillips RL, Fraser GE (1989) Prospective study of         smoking and caffeine in relation to breast cancer risk in young women.
    exogenous hormone use and breast cancer in Seventh-day Adventists.             UK National Case-Control Study Group. Br 1 Cancer 70: 112—119
    Cancer 64: 591—597                                                         29  White E, Malone KE, Weiss NS, Daling IR (1994) Breast cancer among
14 Richardson S, de Vincenzi 1, Pujol H, Gerber M (1989) Alcohol consump           young US women in relation to oral contraceptive use. 1 Nat! Cancer Inst
    tion in a case-control study of breast cancer in southern France. Int 1        86: 505—514
    Cancer44: 84—89                                                            30  Longnecker MP, Newcomb PA, Mittendorf R, Greenberg ER, Clapp RW,
15 Siskind V, Schofield F, Rice D, Bain C (1989) Breast cancer and breast          Bogdan GF, Baron J, MacMahon B, Willett WC (1995) Risk of breast
    feeding: results from an Australian case-control study. Atn 1 Epidemiol        cancer in relation to lifetime alcohol consumption. / Nat! Cancer Inst
    130: 229—236                                                                   87: 923—929
 16 Rosenberg L, Palmer JR, Miller DR, Clarke EA, Shapiro S (1990) A case      31  Longnecker MP, Paganini-Hill It, Ross RK (1995) Lifetime alcohol
    control study of alcoholic beverage consumption and breast cancer. Aîn 1       consumption and bresst cancer risk among postmenopausal women in
    Epideniiol 131: 6—14                                                           Los Angeles. Cancer Epideiniol Biotnarkers Prevent 4: 721 —725
 17 Ewertz M (1991) Alcohol consumption and breast cancer risk in
    Denmark. Cancer Causes and Control 2: 247—252
© 2002 Cancer Research UK                                                                           British Joumal of Cancer (2002)     87(1!), 1234— 245
</pre>

====================================================================== Einde pagina 81 =================================================================

<br><br>====================================================================== Pagina 82 ======================================================================

<pre>                                          Alcohol, tobacco and breast cancer
                        Collaborative Group on Hormonal Factors in Breast Cancer
1244
     32 Primic-Zakelj M, Evstifeeva T, Ravnihar B, Boyle P (1995) Breast cancer    54 Pike MC, Henderson 13E, Krailo MD, Duke A, Roy S (1983) Breast cancer
        risk and oral contraceptive use in Slovenian women aged 25 to 54. Int 1       in young women and use of oral contraceptives: possible modifying effect
        Cancer 62: 414—420                                                            of formulation and age al use. Lancet ii: 926—930
     33 van den Brandt PA, Goldbohm RA, van’t VeerP (1995) Alcohol and             55 Nomura AMY, Hirohata T, Kolonel LN, Hankin JH, Lee 1, Stemmermann
        breast cancer: results from the Netherlands Cohort Study. Am 1 Epidemiol      G (1985) Breast cancer in caucasian and Japanese women in Hawaii. Nat!
         141: 907—915                                                                 Cancer Inst Monogr 69: 191—196
     34 Levi F, Pasche C, Lucchini F, La Vecchia C (1996) Alcohol and breast       56 Hislop TG, Coldman AJ, Elwood JM, Skippen DH, Kan L (1986) Rela
        cancer in the Swiss Canton of Vaud. Eur 1 Cancer 32A: 2108—2113               tionship beween risk factors for breast cancer and hormonal status. Int
     35 Morabia A, Bernstein M, 1-leritier S, Khatchatrian N (1996) Relation of       1 Epidenuiol 15: 469—476
        breast cancer with passive and actïve exposure to tobacco smoke. .4in 1    57 Kay CR, Hannaford PC (1988) Breast cancer and the pil           — a further
        Epidemiol 143: 918—928                                                        report from the Royal College of General Practitioners’ oral contraception
     36 Rossing MA, Stanford JE, Weiss NS, Habel LA (1996) Oral contraceptive         study. Br 1 Cancer 58: 675—680
        use and risk of breast cancer in middle-aged women. Ani 1 Epideinio! 144:  58 Ravnihar B, Primic Zakelj M, Kosmelj K, Stare J (1988) A case-control
         161—164                                                                      study of breast cancer in relation to oral contraceptive use in Slovenia.
     37 Viladiu P, lzquierdo A, de Sanjosé S, Bosch FX (1996) A breast case           Neop!asuna 35: 109—121
        control study in Girona, Spain. Endocrine, familial and lifestyle factors. 59 Lee HP, Gourlev L, Duffy SW, Esteve 1 Lee), Day NE (1992) Risk factors
        Eur J Cancer Procent 5: 329—335                                               for breast cancer by age and menopausa] status: a case-control study in
     38 l3owlin SJ, Leske MC, Varma A, Nasca P, Weinstein A, Caplan E (1997)          Singapore. Cancer Caucses Control 3: 3 13—322
        Breast cancer risk and alcohol consumption: results from a large case      60 Ngelangel GA, Lacaya LB, Cordero C, Laudico AV (1994) Risk factors for
        control study. Int 1 Epideiniol 26: 915—923                                   breast cancer among Filipino women. Phil 1 Internal Mcd 32: 231—236
     39 Clavel-Chapelon F, Dormoy N, Guibout C (1997) Wine, beer and tobacco       61 Hirose K, Tajima K, Hamajima N, lnoue M, Takezaki T, Kuroishi T,
        consumption and the risk of breast cancer: results from a French case         Yoshida M, Tokudome S (1995) A large 5cale, hospital based case-control
        control study. 1 Epidetnio! Biostat 2: 95—104                                 study of risk factors of breast cancer according to menopausal status. Jpn 1
     40 Swanson GA, Coates RJ, Malone KE, Gammon MD, Schoenberg JB,                   Cancer Res 86: 154
        l3rogan DJ, McAdams M, Potischman N, Hoover RN, Brinton LA                 62 Thomas HV, Key TJ, Allen DS, Moore JW, Dowsett M, Fentiman IS,
        (1997) Alcohol consumption and breast cancer risk among women under           Wang DY (1997) Reversal of relation between body mass and endogenous
        age 45 years. Epideniiology 8: 231—237                                        estrogen concentrations with menopausal status. 1 Nat! Cancer Inst 89:
     41 Thomas DB, Gao DL, Self SG, Allison CJ, Tao Y, Mahloch J, Ray R, Qin          396—397
        Q, Presley R, Porter P (1997) Randomized trial of breast self-examination  63 Garrett PA, Hulka BS, Farber RA (2000) Racial differences in the associa
        in Shanghai: methodology and preliminary results. 1 Nat? Cancer Inst 89:      tion of rare HRAS alleles and breast cancer. Anu JEpidentiol 138: 599—599
        355— 365                                                                   64 Collaborative Group on Hormonal Factors in Breast Cancer (1997) Breast
     42 McCredie MRE, Dite GS, Giles GG, Hopper JL (1998) Breast cancer in            cancer and hormone replacement therapy: collaborative reanalysis of data
        Australian women under the age 40. Cancer Causes Control 9: 189— 198          from 51 epidemiological studies of 52705 women with breast cancer and
     43 Ferraroni M, Decarli A, Franceschi S, La Vecchia C (1998) Alcohol              108411 women without breast cancer. Lancet 350: 1047— 1059
        consurnption and risk of breast cancer: a multicentre Italian case-control 65 Collaborative Group on Hormonal Factors in Breasi Cancer (1996) Breast
        study. Enr 1 Cancer 34: 1403— 1409                                            cancer and hormonal contraceptives: further results. Contraception
     44 Enger Sivl, Ross RK, Paganini-HilI A, Longnecker clP, Bernstein L (1999)      54(Suppl 3): IS— 106S
        Alcohol consumption and breast cancer oestrogen and progesterone           66 Peto It, Pike M, Armitage P, l3reslow NE, Cox DR, l-loward SV, Mantel N,
        receptor status. Br 1 Cancer 79: 1308—1314                                    McPherson K, Peto J, Smith PG (1976) Design and analysis of rando
     45 Hopper JE, Chenevix-Trench G, Jolley D, Dite GS, Jenkins MA, Venter           mized clinical trials requiring prolonged observation of each patient. Br
         DJ, McCredie MRE, Giles GG (1999) Design and analysis issues in a            1 Cancer 34: 585—612
        population-based case-control-family study of the genetic epidemiology     67 Easton DF, Peto J, Babiker AGAG (1991) Floating absolute risk: an alter
        of breast cancer, and the Co-operative Family Registry for Breast Cancer       native to relative risk in survival aisd case-control analysis avoiding an
         Families (CFRBCS). A’Ionogr Nat Cancer lust 26: 95—100                       arbitrary reference group. Stat Mcd 10: 1025—1035
     46 Magnusson C, Baron JA, Correis N, Bergstrom R, Adami H-O, Persson 1        68 Doil R, Peto R, Hall E, Wheatley K, Gray It (1994) Mortality in relation
         (1999) Breast cancer risk following long-term oestrogen- and oestrogen       ship to consumption of alcohol: 13 years’ observation of male British
        progestin-replacement therapy. Int 1 Cancer 81: 339—344                       doctors. Br Mcd 1309: 911— 918
     47 Klillion Women Study Collaborative Group (1999) The Million Women          69 Fletcher CM, Peto R, Tinker CM, Speizer FE (1976) The natura? history of
        Study: Design and Characteristics of the Study Population. Breast Cancer      chronic bronchitis and eunphysenua Oxford: Oxford University Press
        Res 1:73—80                                                                70 Parkin DM, Whelan SL, Ferlayl, Raymond L, YoungJ (1997) Cancer Inci
     48 Gao Y-T, Shu XO, Dai Q, Potter J, Brinton L, Wen W, Sellera T, Kushi          dence in Five Continents, Vol VII: Lyon: IARC Scientific Publication
         I.H, Yuen ZX, Bostick R, Jin F, Zheng W (2000) Menstrual, reproductive    71 Doil R, Peto R (1978) Cigarette smoking and bronchial carcinoma: dose
         factors and breast cancer risk in urban Shanghai, People’s Republic of       and time relationships among regular smokers and lifelong non-smokers.
         China. Int 1 Cancer 87: 295—300                                              1 Epidemie! Comun Hea!th 32: 303—3 13
     49 Johnson KC, 1-lu J, Mao Ythe Canadian Cancer Registries Epidemiology       72 Dorgan JF, Baer DJ, Albert PS, Judd JT, Brown ED, Corle DK, Campbell
         Research Group (2000) Passive and acties smoking and breast cancer risk       WS, Hartman TJ, Tejpar AA, Clevidence BA, Giffen GA, Chandler DW,
         in Canada, 1994—97. Cancer Causes Control 11: 211 —221                        Stanczyk FZ, Taylor PR (2001) Serum hormones and the alcohol-breast
     50 Feigelson HS, Calle EE, Robertson AS, Wingo PA, Thun MJ (2001) Alco            cancer association in postmenopausal women. 1 Nat! Cancer Juist 93:
         hol consumption increases the risk of fatal breast cancer. Cancer Causes      7 10—715
         and Control 12: 895—902                                                   73 Thun MJ, Peto R, Lopez AD, Monaco 1H, HenleyJ, Heath Jr CW, DolI R
     51 Kropp S, Becker H, NietersA, Change-ClaudeJ (2001) Lowand Moderate             (1997) Alcohol consumption and mortality among middle-aged and
         alcohol consumption and breast cancer risk by age 50 among women in           elderly US aduits. N Eng! 1 Mcd 337: 1705— 1714
         Germany. A,n 1 Epidetniol 154: 624—634                                    74 IARC monographs on the evaluation of carcinogenic risks to humans
     52 Thomas DB, Noonan EAand the WHO Collaborative Study of Neoplasia               (1988) Alcohol drinking Vol. 44: Lyon, Franco: International Agency for
         and Steroid Contraceptives (1993) Breast cancer and prolonged lactation.      Research on Cancer
         Int 1 Epidemie? 22: 619—626
     53 Vessey MP, McPherson K, Dol? R (1981) Breast cancer and oral contra
         ceptives: findings in Oxford-Family Planning Association contraceptive
         study. Br Mcd 1282: 2093—2094
      Bnosh Joumal of Cancer (2002) 87(1?), 1234       - 1245                                                                      © 2002 Cancer Research UK
</pre>

====================================================================== Einde pagina 82 =================================================================

<br><br>====================================================================== Pagina 83 ======================================================================

<pre>                                                                                    Alcohol, tobacco and breast cancer
                                                                                    Collaborative Group on Hornional Factors in Breast Cancer
                                                                                                                                                                  1245
 APPENDIX 1.— COLLABORATORS (in alphabetical order
                                                                                    Stiudy, NY, USA: RL Hanson, MC Leske, MC Mahoney, PC Nasca, AO
  of institution, study name, or location)                                          Varma, AL Weinstein; Uuiiversity Hospital, Liind, Suvedeui: TR Moller, H
 Aichi Cancer Research Institute, Nagoya, Japan: N Hamajima, K Hirose, K            Olsson, J Ranstam; Maastricht Unwersity, Maastricht, The Nethierla,ids:
 Tajima; Albert Einstein College of Medicine, NY, USA: T Rohan; America,i           RA Goldbohm, PA van den Brandt; University of Phiilippines, Manila,
 Cancer Society, GA, USA: EE Calle, CW Jr Heath; Atlanta, Emory Univer              Philippines: RA Apelo, J Baens, JR de la Cruz, B Javier, LB Lacaya, CA
 sity, GA, USA: RJ Coates, JM Liff, Aviano Cancer Center, Pordenonc,                Ngelangel; Istituto ‘Mario Negri’, Milan, Italy: C La Vecchia, E Negri; Isti
 Italy: R Talamini; Mahidol University, Bangkok, Thailand: N Chantarakul,           tuto Nazionale Tuniori, Divisione di Statistica Medica e Biometria, Milan,
 S Koetsawang, D Rachawat; Breast Tinnor Collaborative Study, Johiis                Italy E Marubini; Istituto di Statistico A’Iedica e Bionietria, Milan, Italy:
 Hopkiiis Uiiiversity, ?clD, USA: A Morabia, L Schuman, W Stewart, M                M Ferraroni; ;Wouitpellicr Cancer Ceuitre & INSERAI, P.lontpellier, Fra,ice:
 Szklo; Unii’crsity of Qiicensland, Brisba,ie, Aiistralia: C ain, F Schofield,      M Gerber, S Richardson, C Segala; Nairobi Centre for Research in Repro
 V Siskind; British Colunibrn Cancer Agency, BC, Canada: P Band, AJ Cold            dii ction, Nairobi, Kenya: D Gatei, P Kenya, A Kungu, JG Mati; Natiouial
 man, RP Gallagher, TG Hislop, P Yang; Cancer Research Center, University           Cancer Institute, MD, USA: LA Brinton, R Hoover, C Schairer; National
  of Hawan, Hawaii, USA: EM Kolonel, AMY Nomura; Canadian Cancer                    Institute of Child Healthi & Hunian Developnient, MD, USA: R Spirtas;
 Registries Epidemiology Research Group, Canada: J Hu, KC Johnson, Y                National Uuiiversity of Singapore, Singapore: HP Lee; The Netherlands Cancer
 Mao; Catahn Institut of Oiicology, Barcelona, Spain: S De Sanjosé; Centers         Institute, Amsterdam, The Nethierlands: MA Rookus, FE van Leeuwen for
 for Disease Control & Prevention, GA, USA: N Lee, P Marchbanks, HW                 the Netherlands Oral Contraceptives and Breast Cancer Study Group;
 Ory, HB Peterson, HG Wilson, PA Wingo; Central Institute of Cancer                 Neiv Jersey State Departnient of Health, NJ, USA: JA Schoenberg; New Southi
 Research, Berlin, Gerniany: K Ebeling, D Kunde, P Nishan; Centre for               Wales Ccmncer Coinicil, Sydney, Australia: M McCredie; Colunibia University
  Genetic Epidetniology, University of Melbourne, Melbourne, Australia: JE          School of Public Healthi, NY, USA: MD Gammon; Ontario Cancer Treatnicut
 Hopper; Channing Laboratory, Brighani and Wonie,i’s Hospita!, Harvard              & Research Foigndation, Ontario, Canada: EA Clarke; Dcpartnient of Public
 Medical School, MA, USA: G Colditz for Nurses’ Health Study Research               Health & Priniars’ Care, Oxford, UK: L Jones, A Neil, M Vessey, D Yeates;
 Group; Chennai Cancer Institute, Madras, India: V Gajalakshmi; Chiaiig             Cancer Research UK Epidenuology Unit, Oxford, UK (Secretariat): P Apple
                                                                                    by, E Banks, V Beral, D Bull, B Crossley, A Goodill, J Green, C Hermon,
                                                                                                                                                                    1
 Îvlai Univcrsity, Chiang Mai, Thailand: N Martin, T Pardthaisong, S Silpi
 sornkosol, C Theetranont; Chulalongkorn University, Bangkok, Thailand: B           T Key, N Langston, C Lewis, G Reeves; Cancer Research UK/MR /BHF
 Boosiri, S Chutivongse, P Jimakorn, P Virutamasen, C Wongsrichanalaj;              Clinical Tricil Service Unit & Epidennobogical Studies Unit, Oxford, UK: R
.Danish Cancer Society, Aalborg, Denniark: M Ewertz; Departnieiit of Medical        Collins, R DoIl, R Peto; Radiation Effects Research Foundatio,i, Hiroshima,
 Epideniiology, Karolinska Institute, Stockhohn, Sweden: HO Adami, L Bergk          Japan: K Mabuchi, D Preston; Royal College of General Practitioners Oral
 vist, C Magnusson, 1 Persson; Deutsches Krebsforschungszentruni,                   Contraception Study, Louido,i, UK: P Hannaford, C Kay; University of Costa
 Heidelbcrg, Gcrniaiiy: 1 Chang-Claude; University of Otago, Dii,iedi,i, New        Rica, San Jose, Costa Rica: L Rosero-Bixby; Shanghai Cancer Institute,
 Zeala,id: C Paul, DCG Skegg, GFS Spears; European Institute of Oncology,           Shanghai, China: YT Gao, F Jin, J-M Yuan; Shanghai 1,istitute of Planned
 Milan, Italy: P Boyle, T Evstifeeva; Fred Hi,tchi,iso,i Caiicer Rc5earch Center,   Parenthiood Research, Shanghai, China: HY Wei, T Yun, C Zhiheng; Dcpart
  WA, USA: JR Daling, WB Hutchinson, K Malone, LA Noonan, JE Stan                   nient of Public Health, Sydney, Austratia: G Berry, 1 Cooper Booth, T
 ford, DB Thomas, NS Weiss, E White; French Multiceiitre Breast St udy,             Jelihovsky, R MacLennan, R Shearman; Tianjin Ccuicer l,istitiute, Tianjin,
 INSERIvI, Villejuijf Fra,ice: N Andrieu, A Brêmond, F Clavel, B Gairard,           China: Q-S Wang; Departnient of Public Healthi Scieuices, Toronto, Canada:
 J Lansac, L Piana, R Renaud; Girojia Caiicer Registry, Girona, Spai,i: A           CJ Baines, AB Miller, C Wall; Troniso Uuiiversity, Troniso, Norivay E Lund,
  lzquierdo, P Viladiu; 1-lospital General de Mexico, Mexico City, Mexico:          II Stalsberg; Vauiderbilt Uuiiversity, TN, USA: XO Shu, W Zheng; Univcrsity
 HR Cuevas, P Ontiveros, A Palet, Sli Salazar; Hospital Universitario, Cali,        of Atheuis Medical School, Atheuis, Greece: K Katsouyanni, A Trichopoulou,
 Colombia: N Aristizabal, A Cuadros; Icelandic Cancer Society, Reykjavik,           D Trichopoulos; University of Cliiie, Santiago, Cliiie: A Dabancens, L
 Iceland: 1. Tryggvadottir, H Tulinius; INSERM, Institiit Gustai’e-Roussey,         Martinez, R Molina, 0 Salas; University of Ediuibiurghi, Edinbiurghi, UK: FE
  Villejuif France: A Bachelot, MG Lê; Institi,te of Cancer Research, Sutton        Alexander; University of Miui,iesota School of Public Healthi, MN, USA: K
 and Londen School of Hygiene and Tropical Medicine, UK: J Peto; Interna            Anderson, AR Folsom on behalf of the lowa Women’s Health Study;
  tional Ageiicy for Research in Cancer, Lyon, France: S Franceschi; Israel          University of North Carohiuia at Chapel Hihl, School of Public Healthi, NC,
  Chaini Sheba Medical Centre, Tel-Hashonier, broei: F Lubin, B Modan, E             USA: BS Hulka; Uuiiversity of Nottinghiauui, Nottinghiani, UK: CED Chilvers;
 Ron, Y Wax; Kaiser Pernianente, A, USA: GD Friedrnan, RA Hiatt; Insti              University of Southern Cahiforuiia, LA, USA: L Bernstein, S Enger, RW
  tut universitaire de niedecuic sociale et prei’entive, Lausapine, Switzerla,id: F Haile, A Paganini-Hili, MC Pike, RK Ross, G Ursin, MC Yu; Unirersity
 Levi; Cancer Research UK Genetic Epideniiology Laboratory, Leeds, UK: T            of Wiscouisin Coniprehensii’e Cancer Center, 1VI, USA: MP Longnecker, P
  Bishop; I,istitute of Oncology, Ljubljana, Slove,iia: K Kosmel), M Primic         Newcomb for the 4 State Study; Vasteras, Sweden: L Bergkvist; World
 Zakelj, B Ravnihar, J Stare; Loma Linda University, CA, USA: WL Beeson,            1-lenitli Organisauion, Gencra, Switzeriand: A Kalache; World Heaith Organi
 G Fraser; Cancer Research UK Departnienf of Matheniatics, Statistics &             sation, UNDP/UNFPA/WHO/World Bank Special Prograuuinie of Research,
 Epidcniiology, Londen: RD Bulbrook, J Cuzick, SW Duffy, IS Fentiman,               Developnient and Research Training iii Human Reprodiuction, Geneva, Swit
 JE Hayward, DY Wang; Loudon School of Hygiene & Tropical Medicine,                 zerland: TMM Farley, S Holck, 0 Meirik.
 Loudon, UK: AJ McMichael, K McPherson; Long Island Breast Cancer
                                                                                                        Analysis and writing committee: Beral V, Bull D, DoIl R,
                                                                                                                                                 Peto R, Reeves G
                                                                                                  Steering committee: Skegg D (Chairman), Colditz G, Hulka 13,
                                                                                                              La Vecchia C, Magnusson C, Muller A, Peterson B,
                                                                                                                             Pike M, Thomas D, Van Leeuven M.
 APPENDIX 11.— References to epidemiological studies of                             APPENDJX 1V. Results on the relation between past
                                                                                                            —
 breast cancer and alcohol and tobacco consumption and to                           smoking and breast cancer in women who reported
 reviews of the topic                                                               drinking no alcohol
  (This cao be viewed on the website www.bjcancer.com)                              (This can be viewed on the website www.bjcancer.com)
 APPENDIX III. Resuits on the relation between current
                        —
  smoking and breast cancer in women who reported
  drinking no alcohol
  (This cao be viewed on the website wws’.bjcancer.com)
  © 2002 Cancer Research UK                                                                                Bntish Joumal of Cancer (2002) 87(11),      234— 1245
</pre>

====================================================================== Einde pagina 83 =================================================================

<br><br>====================================================================== Pagina 84 ======================================================================

<pre>               Springer
  Meta-Analysis of Studies of Alcohol and Breast Cancer with Consideration of the Methodological
  Issues
  Author(s): Tane Key, Susan Hodgson, Rumana Z. Omar, Tina K. Tensen, Simon G. Thompson,
  Alan R. Boobis, Donald S. Davies and Paul Elliott
  Source: Cancer Causes & Control, Vol. 17, No. 6 (Aug., 2006), pp. 759-770
  Published by: Springer
  Stable URL: http://www.jstor.org/stab1e/2973652 1
  Accessed: 23—07—2015 11:28 UTC
  Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at hflp://wwwJstrorgLpag/
  info/aboit/policies/terrnsJsp
  JSTOR is a not-for-profit service that heips scholars, researchers, and students discover, use, and build upon a wide range of content
  in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
  For more information about JSTOR, please contact supportjstor.org.
            Springer- is collaborating with JSTOR to digitize, preserve and extend access to Cancer Causes & Control.
1fs).
flpJLv.jsioLorg
                                     This content downloaded from 145.! 17.203.196 on Thu, 23 lul 2015 11:28:06 UTC
                                                      All use subject to .IS1 t 1 erin nd .‘undiltons
</pre>

====================================================================== Einde pagina 84 =================================================================

<br><br>====================================================================== Pagina 85 ======================================================================

<pre>  Cancer Causes Control (2006) 17:759—770
  DOl 10.1007/s10552-006-001 1-0
   ORIGINAL PAPER
 Meta-analysis of studies of alcohol and breast cancer
 with consideration of the methodological issues
 Jane Key Susan Hodgson Rumana Z. Omar
 Tina K. Jensen Simon G. Thompson
 Alan R. Boobis Donald S. Davies Paul Elliott
 Received: 27 September 2005 / Accepted: 12 January 2006
 © Springer Science+Business Media B.V. 2006
 Abstract                                                                        were robust to study design and analytic approaches in the
 Objective To give an up-to-date assessment of the asso                          meta-analyses. For studies judged high quality, controlled
 ciation of alcohol with female breast cancer, addressing                       for appropriate confounders, excess risk associated with
 methodological issues and shortfalls in previous overviews.                    alcohol drinking was 22% (95% CI: 9—37%); each addi
 Methods Meta-analysis of studies (any language) pro                            tional 10 g ethanollday was associated with risk higher by
 viding original data on incidence of first primary breast                       10% (95% CI: 5—15%). There was no evidence of publi
 cancer and alcohol. Two reviewers independently extracted                      cation bias. Risk did not differ significantly by beverage
 data. Study quality assessed by objective criteria including                   type or menopausal status. Estimated population attribut
 degree of control for confounding; funnel plots examined                       able risks were 1.6 and 6.0% in USA and UK, respectively.
 for publication bias; meta-regression techniques to explore                     Conciusions Taking account of shortcomings in the study
heterogeneity. Risks associated with drinking versus not                        base and methodological concerns, we confirm the alco
drinking and dose—response not constrained through the                          hol—breast cancer association. We compared our results to
origin estimated using random effects methods.                                  those of an individual patient data analysis, with similar
Resuits Ninety-eight unique studies were included,                              findings. We conclude that the association between alcohol
involving 75,728 and 60,653 cases in drinker versus non-                        and breast cancer may be causal.
drinker and dose—response analyses, respectively. Findings
                                                                                Keywords Alcohol Breast
                                                                                cancer Epidemiology Meta-analysis
                                                                                                          .
J. Key• S. Hodgson P. Elliott ()
Department of Epidemiology and Public Health, Faculty of
Medicine, Imperial College London, St Mary’s Caanpus, Norfolk
Place, London, W2 1PG, UK
e-mail: p.elliott@imperial.ac.uk                                                Introduction
R. Z. Omar
                                                                                Meta-analysis provides a succinct and statistically power
Department of Statistical Science, University College London,
London, UK                                                                      ful summary of data from different studies [1]. However,
                                                                                there are particular challenges when meta-analysis is
T. K. Jensen                                                                    applied to observational data, as, unlike randomized con
Department of Environmental Medicine, Institute of Public
                                                                                trolled clinical trials (RCTs), they are prone to confounding
Health, University of Southern Denmark, Odense, Denmark
                                                                                and various biases, which might distort the resuits [2]. We
S. G. Thompson                                                                  explore here the application of meta-analysis to studies of
MRC Biostatistics Unit, Institute of Public Health, Cambridge                   the association of alcohol and breast cancer with particular
                                                                                attention to issues of confounding and bias in observational
A. R. Boobis D. S. Davies
Section of Expenmental Medicine and Toxicology, Imperial                        data. Our aim was to carry Out a more complete assessment
College London, London, UK                                                      of these issues than in previous meta-analyses [3—7] so as
                                                                                                                                       Springer
                                 This content donloaded from 145.117.203.196 on Thu. 23 Jul 2015 11:28:06 UTC
                                                  All use subject to .1 S f( ik 1 çtfl1 and (. ond ii un,
</pre>

====================================================================== Einde pagina 85 =================================================================

<br><br>====================================================================== Pagina 86 ======================================================================

<pre>760                                                                                                 Cancer Causes Control (2006) 17:759—770
to provide robust quantitative estimates of the alcohol—                   panty), family history of breast cancer, socio-economic
breast cancer association to guide public health policy. We                status, oral contraceptive use/hormone replacement ther
focus in particular on issues of study quality including                   apy. Data were abstracted and studies scored independently
treatment of confounders, and on the problems associated                   by two reviewers (JK, SH); any discrepancies were referred
with the reporting and analysis of alcohol consumption,                    to a panel (RO, TJ, PE, ST) for resolution.
with consideration of methods to investigate dose—response                    To avoid violating independence assumptions, studies
and heterogeneity of effect between studies. We compare                    were inciuded once only; for the same reason, only one set
our resuits with those of a recent meta-analysis of mdi                    of controls could be inciuded. We therefore decided, a
vidual patient data (IPD) [8J, which may be less affected by               priori, on the following hierarchy: where a study had been
these problems, and assess the extent to which careful                     published more than once, odds ratios adjusted for the most
application of meta-analysis methods can aid interpretation                appropriate confounders were used in preference; other
and inform policy in an area where RCTs are not feasible.                  wise, the analysis that included the greatest number of
                                                                           participants was used. Where resuits for more than one
                                                                           control group were reported: community was preferred to
Methods                                                                    hospital controls, and non-cancer to cancer controls.
                                                                              Studies were categorized as either retrospective (i.e.,
Studies were identified by searching all relevant databases                case—control or retrospective cohort) or prospective (i.e.,
(Medline, EMBASE, Pascal (BIDS), Science Citation                          follow-up studies, inciuding nested case—control studies).
Index (BIDS), Social Sciences Citation Index (BIDS),                       None of the cohort studies had more than one set of
Index to Scientific and Technical Proceedings (via BIDS),                  con trols.
Biological Abstracts (BIOSIS), Biological Sciences, AIDS
and Cancer Research Abstracts, Biology Digest, Confer                      Statistical analysis
ence Papers Index, Cochrane Library, NHS National
Research Register (NRR), SIGLE (System for Information                     Definition of non-drinker varied between studies and in
on Grey Literature), NTIS (National Technical Information                  some cases included infrequent drinkers (Table 1, studies
Service), TOXLINE) using key words such as breast,                         8, 15, 16, 22, 30, 68, 71, 72), ex-drinkers (studies 1, 3, 14,
neoplasm, and ethanol, and by scanning the references of                   19, 20, 23, 25, 26, 28, 34, 36, 38, 41-43, 51, 52, 54, 57, 59,
identified papers. We used a variety of search methods to                  60, 62, 64, 77—81, 83, 87—89, 91, 93, 95, 98) or both
minimize publication bias, including citation searching,                   infrequent and ex-drinkers (studies 4, 10, 13, 37), while in
identification of grey literature and searches of conference               some studies, the term non-drinker was not further defined
proceedings. The initial search was kept broad in order to                 (studies 2, 5—7, 9, 12, 21, 24, 27, 29, 32, 35, 40, 44, 46, 53,
capture all relevant publications.                                         58, 67, 69, 70, 73, 92, 97, 99). As it was not possible from
    A study was eligible for inciusion if it (i) gave original             the published data to reassign individuals to a common
data, (ii) assessed incidence (not mortality or prevalence),               definition of non-drinker, the study specific definitions
(iii) considered first primary breast cancer, (iv) was pub                 were used, recognizing that this might lead to dilution of
lished in any language between 1 January 1966 and 31                       effect. Similarly, beer, wine and spirits were classified
December 2003. We identified 298 papers for abstraction                    according to definitions used in each publication. Alcohol
of which 187 were excluded because of duplication,                         consumption was converted to g/day using conversion
inappropriate or missing data, or not reporting original                   factors appropriate to each country [9]. As the data on
research (i.e., editorial, comment or review), leaving 111                 alcohol consumption were presented categorically, we used
for inclusion in our meta-analysis. These 111 papers related               the midpoint of each consumption band to estimate dose—
to 98 unique studies.                                                      response, and for the highest consumption band (which was
    We used a simple scoring system to assess study quality                usually open-ended) we assigned a value half the width of
as follows: score 1—studies with inadequate design                         the previous interval above the uppermost cut point [5] (we
(information on alcohol consumption missing for at least                   can-ied Out a sensitivity analysis to this choice).
30% of participants, results not adjusted for age, for case—                  Where estimates of risks were reported for subsets of the
control studies response rate <60%, for cohort studies loss                study population (e.g., pre/postmenopausal, oestrogen
to foHow-up >30%); score 2—studies with acceptable                         receptor status), we used a Woolfe adjusted method [10] to
design but insufficient control for confounding; score                     obtain study-wide risk estimates. We carried out an anal
3—studies with acceptable design and adequate control for                  ysis of drinkers versus non-drinkers with use of random
confounding, defined as control for three or more of the                   effects methods [11] to combine log odds ratios across
following variables: a reproductive characteristic (such as                studies, using a moment estimator of the between study
age at menarche, age at menopause, age at first birth,                     vaniance. Where a study gave a dose—response analysis
    Springer
                              This content downloaded from 145.117.203.196 on Thu, 23 Jul 2015 11:28:06 UTC
                                               All use subjcct to .ISl( )R cr10’, aiid t. ond II it
</pre>

====================================================================== Einde pagina 86 =================================================================

<br><br>====================================================================== Pagina 87 ======================================================================

<pre> Cancer Causes Control (2006) 17:759—770                                                                                                               761
 Table 1 Summary of included studies
 Country                Study II)                   Control Analysis # Cases           # Controls   Confounders
                        and date                                            in most     in most     in most
                                                                            adjusted    adjusted    adjusted
                                                                            analysis    analysis    analysis
 Retrospective studies
   1  Australia         Rohan (1988)                C         E, D            451           451     a, b, c, d, g. h, i, j, k, 1, p, practice of
                                                                                                    breast self-examination
  2   Australia         Price (1999)                C         E, D            276         1,846
  3   Brazil            Gomes (1995)                H         E                144          567
  4   Canada            Rosenberg (1990)            C         E,D             534         1,044     a, b, d, f, g, h, i, j, k, 1, m, p, religion,
                                                                                                    dietary intake, neighbourhood
  5   Canada            Band (2002)                 C         E             1,018         1,025     a, d
  6   Canada            Cotterchio (2003)           C         E,D           2,509        3,511
  7   Canada            Friedenreich (2001)         C         E             1,233         1,237     d
  8   Canada           Lenz (2002)                  H         E,D             556           577     a, b, e, g, h, i, j, 1, m, o, p, age at oophorectomy,
                                                                                                    marital status, proxy respondent status
  9   Chile            Atalah (2000)                H         E                170          340     a
 10   Denmark          Ewertz (1991)                C         E,D           1,361         1.226     a
 11   Finland          Mannisto (2000)             C          E,D             301           443
 12   France           Le (1984)                    H         E,D             500           945     a,b,d,f,g,j,k,l
 13   France           Richardson (1991, 1989)     H          E,D             234           325    a, b, c, d, f, g,j, k, 1, m
 14   France           Viel (1997)                 C          E,D              154          154    a, f, total calorie intalce
 15   Germany          Kropp (2001)                C          E,D             706         1,381    a, d, e, f, j, 1
 16   Gerniany         Nienhaus (2001)             H          E,D             681           651    a, d, j, survey location
 17   Greece           Katsouyanni (1994)          C          E,D             798         1,528    a, b, d, f, g, m, total energy intake, place of birth
 18   Holland          Van’t Veer (1989)           C          E,D             116           161    a, d, f, g,j, 1, m, p, region, season, energy
                                                                                                   percent fat inta]ce
 19   Italy            Talamini (1984)             H          E, D            368           373    a, b, c, f, g, h, i, 1, m, p, marital status,
                                                                                                   food intake
 20 Italy              Ferraroni (1991, 1993)      H          E,D             210           214    a,b,c,d,f,g,j,1,m
21 Italy               La Vecchia (1989),          H          E, D         2,402         2,220     a, b, c, d, f, g, h, i, j, 1, p, geographic area,
                       Soler (1999),                                                               marital status, intake of meat, fats
                       La Vecchia (1985)                                                           and green vegetables
22    Italy            Ferraroni (1998)            H          E, D         2,425         2,437     a, b, f, g, j, 1, m, total energy intake
23    Italy            Toniolo (1989)              C          E, D            250           499    a, d, m, total energy intake
24    Sicily/Italy     Cusimano (1989)             H          E               143           286    a,l
25   Italy             Franceschi (1991)           H          E, D            132           499    a, g, 1, meat and vegetable intalce
26   Japan             Kato (1989)                 H         E              1,740        8,920     a
27   Japan             Hirose (1995),              H          E, D          1,036      20,797      a, d
                       Hirose (2003)
28   Japan             Kikuchi (1990)              C         E                  48           48    a
29   Japan             Kato (1992)                 H          E               899           899
30   Korea             Choi (2003)                 H         E                346           377    a,j
31   New Zealand       Sneyd (1991)                C         E. D             840        1,782     a, b, f, 1, p
32   Nigeria           Adebamowo (1999)            H         E                251           251
33   Poland            Pawlega (1992)              C         E                122           239    a, d, 1, m, p, marital status, no. of persons
                                                                                                   in household
34   Russia            Zarridze (1991)             C                          139           139    a,b,d,g
35   Spain             Viladia (1996)              C         E, D             330           346    a, c, d, g,
36   Spain             Martin-Moreno (1993)        C         E, D             762          988     a, b, c, d, g, j, 1, m, geographic region,
                                                                                                   total energy intalce
37   Sweden            Ranstam (1995)              C                          393          449
38   SwedenlNorway     Adami (1988)                C         E,  D            422          527     a, b, d, f, g, i, j, k, 1, p
39   Switzerland       Levi (1996)                 H         E,  D            230          507     a, c, d, f, g, h, i, j, 1, p, marital status
40   Switzerland       Morabia (1996)              C         E,  D            150           336    a, b, g, i, j, k, 1, m, saturated fat intake
41   UK                Meara (1989)                H         D                998          998     a, b, d, g, i, j, 1, m, p
42   UK                Meara (1989)                C         D                118           118    a,b,d,g,i,j,l,m,p
43   UK                Srnith (1994)               C         E,  D            753          753     a, b, d, e, f, g, i, j, k, p
44   USA               Boice (1995)                C         E,  D            521        2,611     a, b, c, d, f, g, j, k
45   USA               Vachon (2001)               C         E                558        8,744     a, p, birth cohort, familial clustenng,
                                                                                                   source of information
46   USA               Dupont (1989)               H         E                113        2,483     a, length of follow-up
47   USA               Byers (1982)                H         E, D          1,297           751     a
                                                                                                                                                  Springer
                                This content downloaded from 145.117.203.196 on Thu, 23 Jul 2015 11:28:06 UTC
                                                 All usc subject to .1 1 ( >R 1 ernis and Cond dun
</pre>

====================================================================== Einde pagina 87 =================================================================

<br><br>====================================================================== Pagina 88 ======================================================================

<pre>762                                                                                                     Cancer Causes Control (2006) 17:759—770
Table 1 continued
Country              Study ED                  Control Analysis 4 Cases           4 Controls   Confounders
                     and date                                           in most   in most      in ff051
                                                                         adjusted adjusted     adjusted
                                                                         analysis analysis     analysis
 48   USA            Harris (1988)             H          E              1,467    10,178       a
 49   USA            Harvey (1987)             C          E, D           1,524      1,896
 50   USA            O’Connell (1987)          C          E                275      1,519      a
 51   USA            Webster (1983),           C          E, D           1,206      1,256      a, d, g.  j, k, 1, m, p, religion
                     Chu (1989)
 52   USA            Young (1989)              C          E, D             255        358
 53   USA            Nasca (1994, 1990)        C          E, D           1,608      1,609      a, d, g.  j, k, o
 54   USA            Miller (1989)             H          E                404        421      i
 55   USA, Canada,   Enger (1999),             C          E, D           1,844      1,817      a, d, q
      Western Europe Longnecker (1995)
 56 USA              Bowlin (1997)             C          E, D           1,211      1,214      a, b, e, g,j, k, 1, p, religion, marital status,
                                                                                               ever pregnant
 57 USA              Freudenheim (1995)        C          E, D             738        810      a, b, d, g, j, k, 1, m, intake of calories,
                                                                                               and various nutrients and vitarnins
 58   USA            Harris (1992)             H          E, D             604        520      a, b, c, d, e, f, g, i, j, p
 59   USA            Rossing (1996)            C          E, D             537        489      a, d
 60   USA            Longnecker (1995)         C          E, D          6,662      9,163       a, b, f, g, j, k, 1, m
 61   USA            Brinton (1997),           C          E, D           1,579      1,442      a, b, d, f, g, i,j, k, m, o
                     Swanson (1997)
 62 USA              Newcomb (1999)            C          E, D          3,623      3,783       a, d
 63 USA              Baumgartner (1999)        C          E, D             688        804      a, b, d, e, f, g, i, j, k, 1, m, o, p, physical activity,
                                                                                               energy intake, energy adjusted fat intake
 64 USA              Kabat (1997)              C          E, D              42         64      a, f, m, o, p, eostrogen metabolite ratio,
                                                                                               chronic condition
 65 USA              Kinney (2000)             C          E, D             856        784      a, b, f, j, k, 1, m, o, p
 66c USA             Zheng (2003)              H          E, D             317        334      a, d, e, g. j, m
 67 USA              Claus (2001)              C          E                959        986      a, b, c, d, f, g, h, i, j, k, 1, m, o, p, history of
                                                                                               at least one screening mamniogram
                                                                                               one year before interview
 68 USA              Wu (2003)                 C          E                490        591
 69 USA              Zhu (2003)                C          E, D             288        291      a, b, d, f, g, h, j, k, 1, m, p, employment,
                                                                                               marital status, nurnber of people in household,
                                                                                               religion, use of electric blanketimattress pad,
                                                                                               physical activity, on a diet to lose weight,
                                                                                               number of miscarriages, having an infertility test,
                                                                                               intake of vitamins, total energy intake
 70 USA              Gamnson (2002)            C          E             1,508       1,556      a
 71 USA              Li (2003)                 C          E, D             967        998      a, j, m
 72 USA              Wrensch (2003)            C          E, D             285        286      a, b, d, e, f, h, i, j, k, 1, m, p, religion, number of
                                                                                               mammogranis, previous radiation treatment
 73 USA              Xiong (2001)              C          E                100        105
 74 USA!             Rosenberg (1982)          H          E             1,146      2,694       a, c, d, f, g, j, k, 1, n, o, religion, geographic area,
      Canadallsrael                                                                            year of interview, number of previous
                                                                                               hospital admissions
 75 Uruguay          Ronco (1999)              H          E                400        405
 76 b                Royo-Bordonada (1997) C              E, D             315        364      a, b. c, d, f, g, h,  j, k, m, p
 77 Combined         Howe (1991)               C          E, D          1,573       1,974      a, d
      analysis
Prospective studies
 78 Canada           Fnedenreich (1993),                  E, D          1,336      5,238       a, b, d, f, j, practice of breast self-examination,
                     Rohan (2000)                                                              study center, energy intake, study allocation
 79   Denmark        Hoyer (1992)                         D                 51     5,156
 80   Denmark        Tjonneland (2003)         C          E, D             416    23,533       a, f, g, h, k, 1, m
 81   Holland        van den Brandt (1995)                E, D             422      1,579      a, b, c, d, f, g, i,j, k, 1, m, p, energy intake
 82   Sweden         Holmberg (1995)                      E, D             276        452      a, f, g, j, 1, m
 83   Sweden         Lahmann (2003)            C         E,  D             246    11,913
 84   USA            Zhang (1999)                        E,  D             221     2,543       a, c, d, f, g, h, 1, m, p, physical activity index
 85   USA            Zhang (1999)                        E,  D              66     2,218       a, b, c, d, f, h, 1, m, p, physical activity index
    Springer
                             This content downloaded from 145.117.203.196 on Thu, 23 Jul 2015 11:28:06 UTC
                                              All use subject to .1 1 ( ) R erlu snd (.oncl th
</pre>

====================================================================== Einde pagina 88 =================================================================

<br><br>====================================================================== Pagina 89 ======================================================================

<pre>  Cancer Causes Control (2006) 17:759—770                                                                                                                       763
  Table 1 continued
 Country                  Study 11)                        Control Analysis 4 Cases       # Controls    Confounders
                          and date                                               in most  in most       in most
                                                                                 adjusted adjusted      adjusted
                                                                                 analysis analysis      analysis
   86 USA                 Simon (1991)                              E, D             87      1,827      a, b, f, g, j, 1, m, p, subscapular and
                                                                                                        triceps skin foids
   87 USA                 Hiatt (1984)                              E, D           838      87,570      a
   88 USA                 Schatzkin (1987)                          E, D           121       7,067      a, b, d, f, g, j, 1, m, dietaxy fat
   89 USA                 Barrett-Connor (1993)                     E                15        575
   90 USA                 Hiatt (1988)                              D              287      58,044      a, m, o, p
   91 USA                 Zhang (1999),                             E, D         3,483      85,335      a, b, c, d, f, g, h, j, k, m, Iength of follow-up,
                          Willett (1987), Chen (2002)                                                   total energy intake
   92 USA                 Graham (1992)                             D              367       3,670      a, b, d, f, g, h, i, j, k, 1, m, p, fat, tibre and
                                                                                                        energy intake
   93 USA                 Cerhan (1998)                             E               46       1,760
   94 USA                 Lucas (1998)                              E, D           121       7,894      a, b, c, d, f, g, h, j, k, m, p, physical activity
   95 USA                 Potter (1995), Gapstur (1992)             E, D           939      36,166      a, b, f, g, i, j, m, q, type of menopause,
                                                                                                        history of bilateral oophorectomy
   96 USA                 Garland (1999)                            E, D           435    116,236       a, b, d, f, g,j, k, m
   97 USA                 Feigelson (2003)                 C        E, D         1,303      65,258      a, b, c, f, g, h, j, k, 1, m, o, dietary folate,
                                                                                                        methionine, multivitamin use,
                                                                                                        mammographic history, physical activity,
                                                                                                        adult weight gain, energy intake
   98 USA                 Horn-Ross (2002)                 C        E, D           681    104,454       a, b, f, g, j, o, daily caloric intake, physical activity
   99 Western Europe      Clavel-Chapelon (2002)           C        E, D         2,758    276,473       a, b, f, g, 1, m, p, energy intake, follow-up time
 100 Combined             Smith-Wamer (1989)                                                            a, b, d, f, g, h, i, j, k, 1, m, p, fat,
      analysis                                                                                          fibre and energy intake
   There is a small overlap of cases between this study and study 21, it is therefore only used in a sensitivity anlaysis
 b
   From 5 countries—Germany, Switzerland, Northem Ireland, Holland, Spain
   Overlap of cases with 67, used in a sensitivity analysis for “ever never” and in the main analysis for dose—response
 H, case—control study with hospital controls; C, case—control study with community controls; E, “ever” versus “never” drinkers analysis;
 D, dose—response analysis
 Key to confounders: a, age; b, age at menarche; c, age at menopause; d, menopausal status; e, breast feeding; f, panty; g, age at first birth; h, HRT
 use; i, oral contraceptive use; j, family history of breast cancer; k, history of biopsy for benign breast disease; 1, socio-economic status; m, BMI;
n, obesity; o, ethnicity; p, smoking status; q, oestrogen receptor status
A list of references referred to in this table are available from the authors on request
only, we calculated a crude odds ratio of drinkers versus                           results with a model that was constrained to go through the
non-drinkers using the number of cases and controls in                              origin (zero intercept model). Finally we carried out a
each consumption band. This was not possible for eight                              meta-analysis of dose—response slopes using random
studies where either data on number of cases and controls                           effects methods [11].
were not given (Table 1, studies 41, 42, 79, 90, 92) or data                            We carried out a sensitivity analysis to assess how dif
could not otherwise be pooled (studies 20, 34, 37), so these                        fering quality criteria (via the simple scoring system) and
studies were excluded from the drinkers versus non-                                 control for confounding affected the size of the risk esti
drinkers analysis.                                                                  mate, giving seven separate analyses for each question of
    Initial exploration of the dose—response data using the                         interest. We examined possible heterogeneity in results
“pool-first” method [12], which pools study data before                             across studies using the Q statistic [10]. Meta-regression
trend analysis, indicated a monotonic increasing function                           with random effects [13] was used to explore heterogene
relating alcohol consumption with breast cancer risk;                               ity. Characteristics of the studies examined for heteroge
therefore we assumed that the logarithm of the odds ratio                           neity were as follows: whether the data were collected
varied linearly with alcohol consumption. We also tested                            before or after disease onset; for case—control studies
for a quadratic association. We calculated dose—response                            whether the controls were hospital or community based;
slopes (among drinkers) for each study with available data                          pre or postmenopausal; and nationality of the study popu
by use of log linear regression and a variable intercept; that                      lation (USA or CanadalEurope/other).
is, we excluded non-drinkers and hence did not constrain                                Estimates of population attributable risks [14] for the
the curve to go through the origin. We also compared                                USA and UK (calculated as a weighted average of that in
                                                                                                                                                           Springer
                                     This content downloaded from 145.117.203.196 on Fhu, 23 Jul 2015 11:28:06 UTC
                                                      All use subject to .1 S ( )R entts and 1_onti It iun
</pre>

====================================================================== Einde pagina 89 =================================================================

<br><br>====================================================================== Pagina 90 ======================================================================

<pre>764                                                                                                        Cancer Causes Control (2006) 17:759—770
England, Scotland, Wales and Northern Ireland) were                       Fig. 2g). We analysed data separately for drinkers versus
obtained from surveys of drinking habits among women                      non-drinkers of beer (30 studies), wine (32 studies) and
stratified by age [15, 16], by use of age-specific cancer                 spirits (31 studies) where relevant data were available;
registration data for the USA [17] and UK [18], and                       combined least adjusted odds ratios were estimated to be
assuming 12 g of ethanol in an “average” drink in the                      1.16 (95% CI: 1.04, 1.29) for beer, 1.14 (95% CI: 1.05,
USA [9] and 8 g in a unit of alcohol in the UK [19]. We                    1.24) for wine and 1.14 (95% CI: 1.06, 1.23) for spirits.
used percentages of drinkers in each age group (and cate
                                                                          Dose—response
gories of amounts consumed). In the USA, these data were
available for five age groups and five drinking categories
                                                                          Figure 3 is a scatter plot of the log odds ratios of risk of
(no drinks in the past year, 1—11 drinks in the past year, 3
                                                                          breast cancer associated with drinking alcohol. Also shown
drinks per week, 3—7 drinks per week, >7 drinks per week)
                                                                          on the plot are the linear and quadratic fits to the data using
[16]. In the UK, there were seven age groups and five
                                                                          the “pool-first” method. Data are for the least adjusted
drinking categories (<1, 1—7, 7—14, 14—21, >21 units per
                                                                          odds ratios from all studies that provided number of cases
week) [15]. The percentage of drinkers (% heavy drinkers,
                                                                          and controls per drinking category (these data are required
defined as the highest category of alcohol drinking in each
                                                                          for the “pool-first” trend calculation). The quadratic fit is
country) by age were, in the USA: 18—24 years 56% (4%),
                                                                          not significantly better than the linear fit at the 5% level.
25—44 years 66% (4%), 45—64 years 56% (4%), 65—
                                                                                Table 2 and Fig. 4 give results of the meta-analysis of
74 years 41% (3%), 75 years 29% (2%) [16]. In England
                                                                          dose—response and show, amongst drinkers, the higher risk
these percentages were: 15—24 years 79% (17%), 25—
                                                                          associated with drinking an extra 10 g of ethanol a day.
34 years 77% (11%), 35—44 years 77% (10%), 45—
                                                                          Again, results for the seven analyses are shown separately
54 years 74% (10%), 55—64 years 66% (6%), 65—74 years
                                                                          according to degree of control for confounding and study
53% (4%), 75 years 46% (3%) [15]. (Similar data were
                                                                          quality. The combined estimate of excess risk ranged from
available for Scotland [15]; we assumed drinking habits in
                                                                           10% (95% CI: 5, 15%) (multivariate adjustment for con
Wales and Northern Ireland to be the same as in England.)
                                                                          founders in studies with score 3, Fig. 3g) to 13% (95% CI: 9,
The calculation uses resuits from the dose—response anal
                                                                           17%) (least adjusted, studies with score 2 or 3, Fig. 4b).
ysis. All analyses were carried Out using Splus.
                                                                          Prom the studies judged of high quality with control for
                                                                          appropriate confounders (Fig. 4g), and assuming in the USA
Results
Table 1 gives case and control numbers (most completely
adjusted analyses) and brief details of all included studies,
by country and dates of study, for both retrospective and
prospective designs.
                                                                            Û
                                                                             >.
Drinkers versus non-drinkers                                                 D
                                                                            CID
Figure 1 shows crude odds ratios with 95% confidence
intervals for the 89 studies included in the analysis of
drinkers versus non-drinkers. Studies are ranked according
to precision. Overall 29 studies had estimated odds ratio
<1 and 60 studies 1, with combined estimate of 1.11(95%
CI: 1.06—1.17). Table 2 gives resuits, including Q-statis
tics, which give a measure of heterogeneity, of the meta
analysis for seven separate analyses according to degree of                                         0.5       1.5
control for confounding and criteria for study quality                                   0125  0.25    0.67 1     2 4   8  15
(scores of 1, 2 or 3, see Methods). This sensitivity analysis                                                 Estimated Odds Ratio
shows effects of study quality and differing control for                  Fig. 1 Individual study estimates of crude odds ratios (log scale) of
confounding on size of the estimate. The results are also                 the risk of breast cancer associated with drinkers versus non-drinkers
shown in Fig. 2. The estimates ranged from 1.11(95% CI:                   and 95% confidence intervals. The estimates are ranked top to bottom
                                                                          by precision. Area of box showing study point estimate is
1.06—1.17) (least adjusted estimate including all studies,                proportional to precision. Study ID (Table 1) is given down the
Fig. 2a) to 1.22 (95% CI: 1.09—1.37) (multivariate adjust                 left-hand side. The diamond at the bonom of the plot denotes the
ment for confounders in the 19 studies with score 3,                      random effects estimate of the combined result
   Springer
                             This Content downloaded from 145.117.203.196 on Thu, 23 Jul 2015 11:28:06 UTC
                                              All use subject to J 1 (t R 1 er! n, and (ond iii 0
                                                                  -
</pre>

====================================================================== Einde pagina 90 =================================================================

<br><br>====================================================================== Pagina 91 ======================================================================

<pre> Cancer Causes Control (2006) 17:759—770                                                                                                            765
 Table 2 Results of rneta-analysis
 Model           Drinkers versus non-drinkers                                               Dose—response
                 DR (95% CI)                # Studies            Q-Statistic                Percent excess risk (95% CI)     4 Studies       Q-Statistic
                                                                                            per 10 g ethanol per day
 a               1.11 (1.06,  1.17)          89                  319                        12 (9, 15)                       71              124
 b               1.12 (1.06,  1.18)         61                   214                        13 (9, 17)                       54               98
 c               1.17 (1.09,  1.26)         35                   120                        11 (7, 15)                       41               57
 d               1.17 (1.09.  1.26)         28                    76                        12 (8, 17)                       34               52
 e               1.16(1.10,   1.24)         54                   165                        11 (7, 14)                       63              102
 f               1.17 (1.10,  1.24)         42                   106                        12 (8, 16)                       51               91
 g               1.22 (1.09,  1.37)          19                   54                        10 (5, 15)                       33               56
 a, Least adjusted odds ratios from all studies
 b, Least adjusted odds ratios, studies with score 2 or 3
 c, At least age adjusted odds ratios from all studies
 d, At least age adjusted odds ratios, studies with score 2 or 3
 e, Multivariate adjusted odds ratios from all studies
f, Multivariate adjusted odds ratios, studies with score 2 or 3
g, Multivariate adjusted odds ratios, studies with score 3
 an “average” drink contains 12 g of ethanol [9], a woman                               to each study, using the most completely adjusted analyses
drinking an average of two drinks per day compared to a                                 for studies that scored 3, for the variable and zero intercept
 woman who drinks on average one drink per day has a risk                               models for dose—response.
estimated to be 12% (95% CI: 7—19%) higher. For the UK,
 where an “average” drink contains 9.5 g ethanol [9], the                              Heterogeneity
 estimated risk is 10% (95% CI: 5—15%) higher for two drinks
 per day compared with one. Figure 5 shows the slopes fitted                           All analyses showed significant heterogeneity (P<0.05)
                                                                                       across studies in size of association between alcohol con
   Model                                      OR (95% Cl)           # studies          sumption and risk of breast cancer (Q-statistics, Table 2).
                                             1. (1 06, 1.17)           89
                                                                                       Of the various factors entered into meta-regression analy
                                                                                        ses to explore the heterogeneity, in the analyses of drinkers
     b                                       1.12 (1.06, 1.18)        61                versus non-drinkers, retrospective (case—control) studies
     c                                       1.17 (1.09 1,26)         35
                                                                                        with hospital controls were associated with significantly
                                                                                       (P’<0.05) higher odds ratio estimates than those with
    d                                        1.17(1.09, 1.26)         28               community controls (for example, odds ratios of 1.39 (95%
    e                                        1.16(110 1.24)           54
                                                                                       CI: 1.21—1.60) and 1.11 (95% CI: 1.02—1.21), respectively
                                                                                       based on multivariate odds ratios from studies scoring 2 or
                                             1.17 (1.10, 1.24)        42               3); for the dose—response analyses, there were no signifi
                                            1.22 (1.09, 1.37)          19
                                                                                       cant differences between analyses based cm hospital and
    9
                                                                                       community controls (for example, odds ratios of 7% (95%
                                                                                       CI: 2—12%) and 13% (95% CI: 10—17%) per 10 g ethanol,
                                                                                       respectively, based on studies scoring 2 or 3). None of the
                                                                                       other variables examined in meta-regression significantly
        1.0            1.2             1.4             1.6            1.8
                                                                                       reduced the heterogeneity across studies.
                             Estimated Odds Ratio
Fig. 2 Estimates of the combined odds ratio and 95% confidence                         Sensitivity analysis
interval for drinkers versus non-drinkers. Each line corresponds to an
analysis with different inclusion criteria according to study quality
(see Methods) and degree of confounding. Odds ratios combined in                       We checked the sensitivity of our resuits to the dose—
each analysis are (a) least adjusted odds ratios from all studies, (b)                 response calculation; sensitivity to fixing the first and last
least adjusted odds ratios, studies with score 2 or 3, (c) at least age                points of the dose—response in each study (via comparison
adjusted odds ratios from all studies, (d) at least age adjusted odds
ratios, studies with score 2 or 3, (e) multivariate adjusted odds ratios
                                                                                       of zero and variable intercept models and by assigning
from all studies, (f) multivariate adjusted odds ratios, studies with                  different values to the highest consumption band where
score 2 or 3, (g) multivariate adjusted odds ratios, studies with score 3              these were open-ended), and by using binomial logistic
                                                                                                                                               Springer
                                    This content downloaded from 145.117.203.196 on Thu, 23 Jul 2015 11:28:06 UTC
                                                       All use subject to .1 S 1 ( IR 1 erin’, id Ijoid t i ii,
</pre>

====================================================================== Einde pagina 91 =================================================================

<br><br>====================================================================== Pagina 92 ======================================================================

<pre>766                                                                                                                  Cancer Causes Control (2006) 17:759—770
Fig. 3 Scatter plot of the log                       2.5
odds ratios of risk of breast                                                                                      0
cancer associated with alcohol
consumption. Solid and dashed                        2.0
lines show the linear and
quadratic fits, respectively using
the “pool-first’ method. Non-
drinkers are inciuded in the                         1 .5
model fits
                                                                                                                0
                                                 0
                                                     1.0•               0          0
                                                                         0                      0    0
                                                 (1)                      0                                      0
                                                 0                                                                        0
                                                 0
                                                 0             0                       08
                                                                                                   0
                                                                                                        8  0
                                                 0)
                                                 o 0.5
                                                                              0o                             40                     60
                                                                            0     co     0
                                                                                       00             0
                                                                                                         0
                                                                 0
                                                                                              Alcohol consumpion, g/day
rather than log linear regression to estimate the                                    assessment of the association of drinking versus not-
dose—response curve at the study level. We also checked                              drinking alcohol, extensive sensitivity analysis to quality of
sensitivity to alternative choice of controls where these                            included studies and adjustments for confounders, assess
were reported. None of these appreciably altered the                                 ment of the dose—response relationship among drinkers
resuits. As can be seen in Fig. 1 there was no indication                            (i.e., excluding non-drinkers), and exploration of risk by
that smaller studies (indicated by large confidence inter-                           type of alcoholic beverage. We also include an estimate of
vals) were more positive. Formal funnel plots [20] also did                          population attributable risk. Based on these extensive
not indicate any evidence for publication bias, including                            analyses, previous estimates of the positive association of
for subset analyses.                                                                 alcohol to breast cancer are shown to be robust.
Population attributable risk
                                                                                          Model                                        Excess risk (95% Cl)# studies
We estimated the population attributable risk among                                         a                         1                   12%(9%,15%)          71
drinkers of alcohol in the USA and UK to be 1.6 and 6.0%,
respectively. We assessed the sensitivity of these estimates                                b                                             13% (9%, 17%)        54
by recalculating them based on the lower and upper 95%
                                                                                            c                                             11%(7%,15%)          41
confidence interval for the estimated slope; with these
sensitivity limits, our population attributable risk estimate                              d                                              12%(8%,17%)          34
ranged from 0.9 to 2.4% in USA and 3.2 to 8.8% in the UK.
                                                                                           e                                              11%(7%,14%)          63
                                                                                            t                                             12%(8%,16%)          51
Discussion
                                                                                           g                                              10%(5%,15%)          33
This is the largest and most comprehensive meta-analysis
to date ôf the relationship of alcohol to breast cancer, and
in our view provides a sound basis for guiding public
                                                                                      0%          5%             10%        15%          20%
health policy in this area. We included 98 studies and some                                                Estimated higher risk per 109 ethano[fday
20,000 more cases than the largest of the previous meta
                                                                                    Fig. 4 Estimates of the increase in risk of breast cancer amongst
analyses [8]. Compared with previous meta-analyses, all of                          drinkers per 10 g ethanollday. Each line corresponds to an analysis
which reported a positive association of alcohol to breast                          with different inciusion criteria according to study quality (see
cancer [3—8], we included non-English publications, an                              Methods) and degree of confounding. (a)—(g) as in Fig. 2
   Springer
                                   This content downloaded from 145.117.203.196 on Thu, 23 Jul 2015 11:28:06 UTC
                                                     All use subject to .1’ 1 ( )R 1 ern1, and (oud t
</pre>

====================================================================== Einde pagina 92 =================================================================

<br><br>====================================================================== Pagina 93 ======================================================================

<pre>  Cancer Causes Control (2006) 17:759—770                                                                                                   767
                                                                                by comparing analyses of least adjustment, at least age
                                                                                adjustment, and multivariate adjustment. Although resuits
                                                                                varied, positive and significant associations were found in
                                                                                all analyses. Pooling multivariate adjusted results from
                                                                                studies of adequate design with sufficient control for con
                                                                                founding, is likely to be the optimal analysis in terms of
   0                                                                            accounting for bias, assuming the studies are sufficiently
   0
                                                                                homogeneous. 1f the studies are too heterogeneous, then it
                                                                                may be that a pooled risk estimate is inappropriate.
                                                                                   Consideration of study design is important. Case—con
                                                                                trol studies are more prone to bias than cohort studies, in
                                                                                particular with respect to exposure assessment and recall
                                                                                bias. Among case—control studies, controls are either hos
                                                                               pital or community-based. Ideally controls should be
       Alcohol consmption, g/day          Alcohol consmption, g/day
                                                                                selected independently of exposure, buL hospital patients
 Fig. 5 Comparison of variable and zero intercept models (dose—                 may not be representative of the exposure distnbution in
response). Fitted slopes for most adjusted odds ratios from studies
                                                                                the source population (though authors using hospital-based
scoring 3. Left figure shows the slopes fitted using the variable
intercept model (non-drinkers excluded), nght figure shows the slopes          controls generally stated that they attempted to exclude
fitted using the zero intercept model (non-drinkers inciuded)                   subjects with diseases related to alcohol consumption). We
                                                                               used meta-regression to explore heterogeneity due to these
Methodological considerations                                                   factors. We did not find a significant difference between
                                                                               risks estimated using case—control and cohort studies.
Meta-analysis of observational studies presents challenging                    However, we did find that among case—control studies, risk
 methodological issues involving different study designs                       estimated using hospital-based controls was significantly
 (case—control and cohort), and variation in the quality of                    higher than that using community-based controls for the
 studies in terms of response rate, missing data, exposure                     drinker versus non-drinker analysis—though a significant
 assessment, control for confounding, and choice of controls                   positive association was stijl found after exclusion of
 in the case—control studies. A potential limitation of our                    studies using hospital-based controls—but not for the
 study, in common with other meta-analyses of observa                          dose—response analysis. We also explored for heterogene
 tional studies (compared with meta-analyses of RCTs), is                      ity according to pre/postmenopausal status and country.
 the issue of bias. We made every effort to identify and deal                  Again using meta-regression, we did not find any signifi
 with sources of bias in the published daLa. We carried Out a                  cant differences. It is reassuring to note that in every
 sensitivity analysis based on pre-defined quality criteria                    analysis carried out there was a significant positive asso
and control for confounding. Our scoring system was                            ciation, and therefore the findings are consistent over a
simple and objective, as an over-complicated system might                      range of scenarios.
introduce subjectivity and bias into the analysis. Inclusion                       Bias can be introduced into a meta-analysis by inciuding
of quality scores as covariates or weights in a meta-analysis                  studies favouring a positive result (publication bias) or by
will lead to bias [21]. However, our score criteria were                       abstracting incorrect data. To ensure that publication bias
selected to identify studies with potential biases due to                      was minimized we undertook an extensive literature search
design issues or confounding, so that sucli studies could be                   that was not restricted to publications in English and
excluded in the sensitivity analyses (even though this will                    included searching grey literature; we found no evidence of
result in increased variance due to smaller sample size).                      publication bias in our analysis. Two researchers mde
We did not include the scores as part of the regression                        pendently abstracted all data and resolved any discrepan
analyses, or as weights.                                                       cies by consensus to reduce observer bias.
    Another possible limitation of meta-analysis of pub                            Misclassification of exposure is another source of bias.
lished data is the issue of confounding as studies differ in                   There is potential for bias if light, infrequent or ex-drinkers
their definitions of the various confounding variables, and                    are classifled as non-drinkers, as was the case in many
the confounders included vary between studies. While our                       studies analysed. However, this bias is not present in our
definition of “sufficient control for confounding” was                         analysis of dose—response since non-drinkers were
broad enough to encompass a range of potential con                             excluded (affecting the vertical placement of the slope buL
founders, it was still able to identify a subset of studies                    not its estimate). In addition, people may under-report the
with at least a similar approach to the treatment of con                       amount of alcohol consumed, especially heavy drinkers
founding. We further explored the effects of confounding                       [22]. In an analysis of dose—response such a bias may
                                                                                                                                       Springer
                                  This content downloaded from 145117.203l96on Thu. 23 Jul 2015 11:28:06 UTC
                                                   All usc subject to 1Sf ( )R t tni and ( oitd t t i t—.
</pre>

====================================================================== Einde pagina 93 =================================================================

<br><br>====================================================================== Pagina 94 ======================================================================

<pre>  768                                                                                                   Cancer Causes Control (2006) 17:759—770
  exaggerate the slope, but should not generate a non-zero                          Not all of the data and analytical problems associated
  slope where there is no association.                                           with meta-analysis can be solved by carrying out an mdi
     An important methodological feature was our use of a                        vidual patient data analysis. For example, the definition of
  variable intercept model when assessing the dose—response                      a non-drinker was not consistent across studies, and
 relationship. There are several reasons for doing this: (i) it                  sometimes included infrequent or ex-drinkers, often
 does not assume that any linear dose—response relationship                      reflecting data captured in the original study questionnaire.
 passes through the origin. For example, at small doses the                      Study design issues such as low response rate or selection
 relationship may be non-linear e.g., with lower risks than                      of controls are also problems that cannot be solved by an
 for zero exposure. Thus a variable intercept model allows                       individual patient data analysis.
 for departure from linearity around the origin, while still                        The Oxford collaborative study [8] is the largest of the
 allowing a linear relationship with doses away from zero;                       previous meta-analyses and included re-analysis of mdi
 (ii) as noted, the reference group (non-drinkers) may be                        vidual data. They were able to include data from 19
 contaminated to some extent by the inciusion of                                 unpublished studies, which were therefore not included in
 ex-drinkes or women who drink only occasionally, which                          our analysis. However, they did not inciude data from 67
 makes it more difficult to estimate the effect around the                       studies, involving over 40,000 cases, which have been
 origin; (iii) to take account of systematic differences (other                  inciuded in the meta-analysis reported here. The Oxford
 than alcohol intake) between women who drink and those                          study did not account for quality of included studies and
 who abstain from alcohol, as this may induce an “appar                          included non-drinkers in their estimate of dose—response.
 ent’’ effect associated with drinking. (iv) 1f there were a                    Despite these differences, results are comparable with ours,
 threshold effect at a low dose of alcohol, a zero intercept                     with the Oxford study finding a 7.1% higher risk for each
 model would induce a dose—response relationship whereas                        additional 10 g ethanol per day compared with our estimate
 a variable intercept model would not. By anchoring all                         of 10% (95% CI: 5—15%) based on studies judged of high
 slopes at the same point, the zero intercept model forces the                  quality with appropriate control for confounding.
 dose—response slopes of each study (i.e., the observed
 relationship) to differ, whereas the variable intercept model                  Biological plausibility
 is more accepting of a common relationship, seen as
 parallel slopes. Therefore, with respect to the estimated                      Given the positive association of alcohol intake to breast
 dose—response slope, the zero intercept model forces more                      cancer is robust and not readily explained by bias, con
heterogeneity between the studies.                                              founding or heterogeneity, a causal interpretation needs to
     Our estimate of population attributable risk in the UK                     be considered. What then, might be the biological mech
 was higher than in the USA, reflecting different drinking                      anism? Whilst alcohol may be directly carcinogenic to the
habits in the two countries; the USA estimate was similar                       breast, it is more likely to act indirectly through one or
to that reported previously [23].                                               more mechanisms. For example, it may influence the
                                                                                metabolism of mammary carcinogens through induction or
Comparison with an individual patient                                           inhibition of P450 enzymes [24, 25]. However, direct
data analysis (IPD)                                                             evidence for such involvement in breast cancer is lacking
                                                                                [26—28].
An individual patient data analysis, where source data are                          Several studies [29—3 1] have reported that alcohol
obtained from the investigators rather than relying on pub                      consumption is associated with an increased amount of
lished accounts, should give a more comprehensive assess                        mammographically dense tissue in the breast. It has been
ment of risk than a standard meta-analysis, particularly with                   found that mammographic density is positively associated
respect to exposure classification and dealing with con                         with plasma insulin-like growth factor 1 (IGF-I) levels and
founders. However, ]PDs are not widely carried Out because                      inversely associated with plasma IGF binding protein 3
they are expensive and time-consuming. Both a standard                          (IGFBP-3) in premenopausal women [32]. Yu and Berkel
meta-analysis and an IPD require data from all relevant                         [33] reported that moderate consumption of alcohol
studies, both published and unpublished, to avoid bias. In                      increases the production of IGFs by the liver and suggested
practice, sample data are unlikely to be available from all                     that elevated circulating levels of IGFs may stimulate or
investigators, and thus, unlike a standard meta-analysis, an                   promote the development andlor growth of breast cancer.
IPD analysis may not include all the published studies. On                          As breast cancer has a hormonal aetiology [34], any
the other hand, they are likely to inciude unpublished data,                    effects of alcohol on the endogenous hormonal milieu in
and inclusion of these data in an TPD analysis may give an                      women could provide a potential mechanism for carcino
advantage over standard meta-analysis.                                          genesis. Alcohol increases endogenous oestrogen levels in
    Springer
                               This content downloaded from 145.117.203.196 on Thu, 23 iii! 2015 11:28:06 UTC
                                                All use subject to ,1 S 1 ( )R 1 ern1 :ind (.ond ii ii
</pre>

====================================================================== Einde pagina 94 =================================================================

<br><br>====================================================================== Pagina 95 ======================================================================

<pre> Cancer Causes Control (2006) 17:759—770                                                                                                               769
pre- and postmenopausal women [35, 36], possibly via an                             8. Collaborative Group on Hormonal Factors in Breast Cancer
increased rate of aromatization of testosterone or decreased                            (2002) Alcohol, tobacco and breast cancer         —  collaborative
                                                                                        reanalysis of individual data from 53 epidemiological studies,
rate of oxidation of oestradiol to oestrone [37], and ele                               including 58,515 women with breast cancer and 95,067 women
vated levels of oestrone suiphate, a long-term indicator of                             without the disease. Br J Cancer 87:1234—1245
oestrogen levels, have been demonstrated in women who                              9. Turner C (1990) How much alcohol is in a “standard drink”? An
regularly consume alcohol [38].                                                         analysis of 125 studies. Brit J Addiction 85:1171—1175
                                                                                  10. Dersimonian R, Laird N (1986) Meta-analysis in clinical trials.
     There were insufficient data in our study to investigate                           Control Clin Trials 7:177—188
possible interactions with hormone replacernent therapy                           11. Hedges LV, 01km 1(1985) Statistical methods for meta-analysis.
(HRT) and with oestrogen receptor/progesterone receptor                                 Academie Press, Orlando, FL
(ERJPR) status of the tumour. More studies are needed to                          12. Greenland S, Longnecker MP (1992) Methods for trend estima
                                                                                        tion from summarized dose—response data, with applications to
assess such possibilities.                                                              meta-analysis. Am J Epiderniol 135:1301—1309
                                                                                  13. Thompson SG, Sharp SJ (1999) Explaining heterogeneity in
Summary                                                                                 meta-analysis: a comparison of methods. Stats Mcd 18:2693—
                                                                                        2709
                                                                                  14. Hanley JA (2001) A heuristic approach to the formulas for
To summarize, we have shown that the epidemiological                                    population attributable risk. 3 Epidemiol Comrnunity Health
evidence of a positive association between alcohol con                                 55:508—514
sumption and risk of breast cancer is robust to the quality                       15. http://www.doh.gov.uk/stats/tables/atrend98 11 .xls, accessed Jan
and type of study included, and cannot readily be explained                            2003 http://www.show.scot.nhs.uk/scottishhealthsurvey/sh809—
                                                                                       04.html, accessed Jan 2003
by bias or confounding. We have compared our resuits with                         16. Schoenborn CA, Dams PF (2001) Alcohol use among adults:
those of an analysis of individual patient data, with similar                          United States, 1997—98. Advance data from vital and health
findings from the two approaches. Although the excess risk                             statistics; No. 324. National Center for Health Statistics,
associated with drinking alcohol is relatively small                                   Hyattsville, Maryland
                                                                                  17. SEER Cancer Statistics Review 1973—1994. National Cancer
compared with the major risk factors for breast cancer [39],                           Institute
it is one of the few modifiable risk factors associated with                      18. Cancer statistic registrations for England, 1998, http://www.
breast cancer. Given the high prevalence of drinking, even a                           wales.nhs.uk/sites/documents/242/Cancerlncidence1992—200 1 .pdf,
small risk linking breast cancer with alcohol, if causal, bas                          accessed Jan 2003 http://www.show.scot.nhs.uk/isd/cancer/excel/
                                                                                       female._breastlb.xls, accessed Jan 2003 http://researchservic
serious public health implications in terrns of the number of                          es 1 .qub.ac.ukinicr/incidencesearch.htm, accessed Jan 2003
breast cancer cases attributable to drinking alcohol.                             19. Hedges B, di Salvo P (1998) Alcohol consumption and smoking.
                                                                                       In: Prescott-Clarke P, Primatesta P (eds) The health survey for
                                                                                       England 1996. The Stationery Office, London
Acknowledgments Study funded by the Department of Health in                      20. Light RJ, Pillemar DB (1984) Summing up: the science of
England. Authors gratefully acknowledge advice and comments from                       reviewing research. Harvard University Press, Cambridge, MA
members and secretariat of the Committee on Carcinogenicity of the               21. Greenland S, O’Rourke K (2001) On the bias produced by quality
Department of Health. Gram support for this work was by UK                             scores in meta-analysis, and a hierarchical view of proposed
Department of Health (META-ANLAYSIS/99).                                               solutions. Biostatistics 2:463—471
                                                                                 22. Rehm J (1998) Measuring quantity, frequency and volume of
                                                                                       drinking. Alcoholism. Clin Exp Res 22:4S—14S
                                                                                 23. Tseng M. Weinberg CR, Umbach DM, Longnecker MP (1999)
References                                                                             Calculation of population attributable risk for alcohol and breast
                                                                                       cancer. Cancer Cause Control 10:119—123
  1. Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F (2000)                   24. Niemela 0, Parkkila S, Juvonen RO, Viitala K, Gelboin HV,
     Methods for meta-analysis in medical research. Wiley                              Pasanen M (2000) Cytochromes P450 2A6, 2E1, and 3A and
 2. Egger M, Schneider M, Davey Smith G (1998) Spurious preci                          production of protein—aldehyde adducts in the liver of patients
     sion? Meta-analysis of observational studies. BMJ 316:140—144                     with alcoholic and non-alcoholic liver diseases. J Hepatol
 3. Longnecker MP, Berlin JA, Orza MJ, Chalmers TC (1988) A                            33:893—901
     nieta-analysis of alcohol consumption in relation to risk of breast         25. Kedderis GL, Batra R, Koop DR (1993) Epoxidation of acrylo
     cancer. JAMA 260:652—656                                                          nitrile by rat and human cytochromes P450. Chem Res Toxicol
 4. Longnecker MP (1994) Alcoholic beverage consumption in                             6:866—871
     relation to risk of breast cancer: meta-analysis and review. Cancer         26. Swann PF (1984) Effect of ethanol on nitrosamine metabolism
     Cause Control 5:73—82                                                             and distribution. Implications for the role of nitrosamines in
 5. Roth HD, Levy PS, Shi L, Post E (1994) Alcoholic beverages and                     human cancer and for the influence of alcohol consumption on
     breast cancer: some observations on published case—control                        cancer incidence. IARC Sci PubI 57:501—512
     studies. J Clin Epiderniol 47:207—216                                       27. Dunn SR, Simenhoff ML, Lele PS, Goyal S, Pensabene JW,
 6. Ellison RC, Zhang Y, McLennan CE, Rothman KJ (2001)                                Fiddier W (1990) N-nitrosodimethylamine blood levels in
     Exploring the relation of alcohol consumption to risk of breast                   patients with chronic renal failure: modulation of levels by
     cancer. Am J Epidemiol 154:740—747                                                ethanol and ascorbic acid. J NatI Cancer Inst 82:783—787
 7. Bagnardi V, Blangiardo M, La Vecchia C, Corrao G (2001) A                    28. Chhabra SK, Souliotis VL, Kyrtopoulos SA, Anderson LM
     meta-analysis of alcohol drinking and cancer risk. Br J Cancer                    (1996) Nitrosamines, alcohol, and gastrointestinal tract cancer:
     85:1700—1705                                                                      recent epidemiology and experimentation. In Vivo 10:265—284
                                                                                                                                                  Springer
                                     This content downloadcd from 145.117.203.196 on Thu. 23 Jul 2015 11:28:06 UTC
                                                     All usc subject to .1 t ( R 1 erms and 1_ond ion
                                                                          ‘
</pre>

====================================================================== Einde pagina 95 =================================================================

<br><br>====================================================================== Pagina 96 ======================================================================

<pre>770                                                                                                             Cancer Causes Control (2006) 17:759—770
29. Funkhouser E, Waterbor JW, Cole P, Rubin E (1993) Mammo                     35. Reichman ME, Judd IT, Longcope C, Schatzldn A (1993) Effects
    graphic patterns and breast cancer risk factors among women                       of alcohol consumption on plasma and urinary hormone con
    having elective screening. South Med J 86:177—180                                 centrations in premenopausa] women. J NatI Cancer Inst 85:722—
30. Vachon CM, Kushi LH, Cerhan IR, Kuni CC, Sellers TA (2000)                        727
    Association of diet and manimographic breast density in the                 36. Dorgan 1F, Baer DJ, Albert PS et al. (2001) Serum hormones and
    Minnesota breast cancer family cohort. Cancer Epidemiol Bio                       the alcohol—breast cancer association in postmenopausal women.
    markers Prev 9:151—160                                                            J Nat) Cancer Inst 93:710—715
31. Vachon CM, Kuni CC, Anderson K, Anderson VE, Sellers TA                      37. Gul J (2000) The effects of moderate alcohol consumption on
    (2000) Association of mammographically defined percent breast                     female bormone levels and reproductive function. Alcohol
    density with epidemiologic risk factors for breast cancer (United                 35:417—423
    States). Cancer Cause Control 11:653—662                                    38. Hankinson SE, WilIeu WC, Manson JE et al. (1995) Alcohol,
32. Byrne C, Colditz GA, Willett WC, Speizer FE, Poflak M,                            height, and adiposity in relation to estrogen and prolactin levels
    Hankinson SE (2000) Plasma insulin-like growth factor (IGF) 1,                    in postmenopausal women. 1 Nati Cancer Inst 87:1297—1302
    IGF-binding protein 3, and mammographic density. Cancer Res                 39. Henderson BE, Pike MC, Bernstein L, Ross RK (1996) Breast
    60:3744—3748                                                                      Cancer. In: Schottenfeld D, Fraumeni 1F Ir (eds) Cancer epide
33. Yu H, Berkel J (1999) Do insulin-like growth factors mediate the                  miology and prevention, 2nd edn. Oxford University Press,
    effect of alcohol on breast cancer risk? Med Hypotheses 52:49 1—                  pp 1022—1039
    496
34. Key TJA, Pike MC (1988) The role of oestrogen and progesto
    gens in the epidemiology and prevention of breast cancer. Eur 1
    Cancer Clin Oncol 24:29—43
   Springer
                                  Thus Content downloadcd from 145.117.203.196        on Thu. 23 Jul 2015    11:28:06 UTC
                                                   All use subject to .1 S It R 1 cri t :111(1 t oncl t i in
</pre>

====================================================================== Einde pagina 96 =================================================================

<br><br>====================================================================== Pagina 97 ======================================================================

<pre>World SN American
Cancer N 4 Institute for

Research Fund Sy Cancer Research

Continuous Update Project
Keeping the science current

Breast Cancer
2010 Report

Food, Nutrition, Physical Activity,
and the Prevention of Breast Cancer

Continuous
Update Project

° P/

</pre>

====================================================================== Einde pagina 97 =================================================================

<br><br>====================================================================== Pagina 98 ======================================================================

<pre>WORLD CANCER RESEARCH FUND GLOBAL NETWORK
OUR VISION
The World Cancer Research Fund global network heips people make choices
that reduce their chances of developing cancer.
OUR HERITAGE
We were the first cancer charity:
•  To create awareness of the relationship between diet and cancer risk
•  To focus funding on research into diet and cancer prevention
•    To consolidate and interpret global research to create a practical
     message on cancer prevention
OUR MISSION
Today the World Cancer Research Fund global network continues:
•    Funding research on the relationship of nutrition, physical activity and
     weight management to cancer risk
•    Interpreting the accumulated scientific literature in the field
•    Educating people about choices they can make to reduce their chances
     of developing cancer
THE WCRF GLOBAL NETWORK
The World Cancer Research Fund (WCRF) global network corn prises WCRF
International, which operates as the umbrella association for the global
network’s four charitable organisations: The Arnerican Institute for Cancer
Research (AICR); World Cancer Research Fund (WCRF UK); World Cancer
Research Fund Netherlands (WCRF NL); World Cancer Research Fund Hong
Kong (WCRF HK).
</pre>

====================================================================== Einde pagina 98 =================================================================

<br><br>====================================================================== Pagina 99 ======================================================================

<pre>Please cite the Report as follows:
World Cancer Research Fund / American Institute for Cancer Research. Continuous
Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of Breast
Cancer. 2010
This Report provides an updated version of section 7.10 Breast Cancer from the Second
Expert Report: Food, Nutrition, Physical Activity and the Prevention of Cancer: a Global
Perspective. This section has been updated with the latest information from the 2008
Continuous Update Project Breast Cancer SLR prepared by a team at Imperial College
London, UK (see acknowledgements). For further details on the epidemiological evidence
please see the full 2008 Continuous Update Project Breast Cancer SLR (Second Expert
Report). For further details on mechanisms please see the            Expert Report.
The First and Second Expert Reports represent the most extensive analysis of the existing
science on the subject to date. To keep the evidence current and updated into the future,
WCRF/AICR is undertaking the Continuous Update Project, in collaboration with Imperial
College London. The Continuous Update Project builds upon the work conducted for the
Second Expert Report and began by merging all the databases from the different cancer
sites into an upgraded database.
The Continuous Update Project provides the scientific community with a comprehensive
and up to date depiction of scientific developments on the relationship between diet,
physical activity, obesity and cancer. It also provides an impartial analysis and
interpretation of the data as a basis for reviewing and where necessary revising
WCRF/AICR’s cancer prevention recommendations based on the 2007 Expert Report.
In the same way that the Second Expert Report was informed by a process of systematic
literature reviews (SLRs), the Continuous Update Project systematically reviews the
science. WCRF/AICR has convened a panel of experts (the Continuous Update Project
Panel (see acknowledgements) consisting of leading scientists in the field of diet,
physical activity, obesity and cancer who consider the evidence produced by the
systematic literature reviews and meta-analyses, and consider the results and draw
conclusions before making recommendations.
The updates to the SLRs are being conducted by a team of scientists at Imperial College
London in liaison with the SLR centres where possible.
Instead of periodically repeating the extensive task of conducting multiple systematic
literature reviews that cover a long period of time, the continuous review process is based
on a live system of scientific data that is updated on an ongoing basis from which, at any
point in time, the most current review and meta-analysis of scientific data can be
pe rform ed.
Periodically WCRF/AICR will produce reports which will outline the scientific
developments in the field of diet, physical activity, obesity and cancer. The reports may
also include updates to the WCRF/AICR recommendations.
The updated recommendations will be used by the WCRF/AICR education and media
relation departments to inform the general public both of the benefits of a healthy
lifestyle and of the developments in science that underpin these recommendations.
</pre>

====================================================================== Einde pagina 99 =================================================================

<br><br>====================================================================== Pagina 100 ======================================================================

<pre>New information in this report
Section 1. Updated with recent mortality and survival data.
Section 2. Updated section on family history
Section 3. No update
Section 4. No update
Section 5. A new section briefly describing the methodology of the Continuous Update
           Project
Section 6. Evidence has been updated based on the 2008 Continuous Update Project
           Breast Cancer SLR and judgements from the Continuous Update Project
           Panel
Section 7. Provides a comparison with the Second Expert Report.
Since publication of this report in 2011, some changes have been
made to the design and formatting, but no changes have been
made to the content of the report or Panel conclusions. Please
note, however, that the Second Expert Report matrix in this report
has been replaced with the Continuous Update Project Matrix (on
page 3).
                                                                                    2
</pre>

====================================================================== Einde pagina 100 =================================================================

<br><br>====================================================================== Pagina 101 ======================================================================

<pre> FOOD, NIJTHITION, PHYSICAL ACTIVITY AND
 I3REAST CANCER (PREMENOPAUSE) 2010
                          OECREASES RISK                        INCREASES RISK
                         bctation                        McGlo4ic drinks
                         Body fati,es                    MuIt attained height’
                                                         reater brtIi weiht
                         Physical activity
                         Diet2r Ibre:                   ar. fruit; scs ad
                         prDJC[a: meit; h; milk. an. dairy prcd .ic: al
                        fai.; fj i:c: vtanir i; :alc r; .3nic index; dicr
                         atteïs; adult aihL 1e r,; ehi’nire atres
                         Nor.e idnLiFd
1 .tLll     r         ..fll.zllIa1Ft mcJ ‘ 1fl:         cn:. 11 a rn.r :l F:n:l,
  c7Vrmcnt4, rmcril.         a : ri-1 ;rlY                p’1h ui lt :d tn
  c.,a!tEr In :ir’ Ln Irci                   :: :h1r 2 1 ::ç2 prl lcçai.
2             cl ni        WnllZç, c.tnchd,         izci rd :c1Ii
 FOOD, NLfl1IITION, PHYSICAL ACTIVFIT AND
 BREAST CANCER (POSTMENOPAUSE) 2010
                         DECREASES RISK                         INCREASES RISK
                         Lactatioiï                      icoiatic dliakl
                                                         Body fatnes
                                                         td uit attained Ieigtt
                         Piiysieal aetivity              Ab4lOrIliTîai fatns
                                                         A41u11 weigtit gain
                                                        Total lat
                         Dieter fbe; ve’eLt:            erd ftt.iLe; eyn and
                         pr:ducs; meat s; rr k nr dairy prduc; olate;
                        vitanin D; caloiLn; ealeriLn: Iycaorr c rdex: d eary
                         pe:erns; birJ, e ht; onery inLae
                         Nc’ie dort. Eed
1       Ie          I Lr       .l::Ly Ic -.xiI l[ rl cl c-a t Is mirr ‘u’ t.l:,
                                                                             1
  nr-I.l-Er-vl j-,:               nj1cç.i                                   odncm
  pxn:clIr Inc:çn:4n, cl mci 1cw ncc cpL C.’
2 P4ca cLry al cdi         nccLij1tcncd, a.e;h4d Irn-.4i -d r:crcaLcni
                                                                                  3
</pre>

====================================================================== Einde pagina 101 =================================================================

<br><br>====================================================================== Pagina 102 ======================================================================

<pre>Cancer of the breast is the most common cancer in women worldwide. Around 1.1 million
cases were recorded in 2004.
Observed rates of this cancer increase with industrialisation and urbanisation, and also
with facilities for early detection. It remains much more common in high-income countries
but is now increasing rapidly in middle- and low-income countries, inciuding within Africa,
much of Asia, and Latin America. Breast cancer is fatal in under half of all cases and is
the leading cause of death from cancer in women (fifth for men and women combined),
accounting for 16 per cent of all cancer deaths worldwide in 2004.
Breast cancer is hormone related, and the factors that modify risk of this cancer when
diagnosed premenopausally and when diagnosed postmenopausally (much more
corn mon) are not the same.
The Continuous Update Project Panel judges as follows:
The evidence that lactation protects against breast cancer at all ages is
convincing.
Physical activity probably protects against breast cancer postmenopause,
and there is limited evidence suggesting that it protects against this
cancer diagnosed premenopause. The evidence that alcoholic drinks are a
cause of breast cancer at all ages is convincing. The evidence that the
factors that lead to greater adult attained height, or its consequences, are
a cause of postmenopausal breast cancer is convincing, and these are
probably also a cause of breast cancer diagnosed premenopause.
The factors that lead to greater birth weight, or its consequences, are
probably a cause of breast cancer diagnosed premenopause. Adult weight
gain is probably a cause of postmenopausal breast cancer. The evidence
that body fatness is a cause of postmenopausal breast cancer is
convincing, and abdominal body fatness is probably also a cause. 0fl the
other hand, body fatness probably protects against breast cancer
diagnosed premenopause. There is limited evidence suggesting that total
dietary fat is a cause of postmenopausal breast cancer.
Life events that protect against breast cancer include late menarche, early
pregnancy, bearing children, and early menopause, all of which have the
effect of reducing the number of menstrual cycles, and therefore lifetime
exposure to oestrogen. The reverse also applies.
See chapter 8 of the Second Expert Report for evidence and judgements
on factors that modify risk of body fatness and abdominal fatness,
including physical activity and sedentary ways of life, the energy density of
foods and drinks, and breastfeeding.
In final summary, the strongest evidence, corresponding to judgements of
“convincing” and “probable” show that lactation protects against breast
cancer; that alcoholic drinks are a cause of this cancer; that the factors
that lead to a greater adult attained height, or its consequences, are a
cause of postmenopausal and probably also premenopausal breast cancer;
that factors leading to greater birth weight, or its consequences, are
                                                                                          4
</pre>

====================================================================== Einde pagina 102 =================================================================

<br><br>====================================================================== Pagina 103 ======================================================================

<pre>probably a cause of premenopausal              breast cancer; and that abdominal
body fatness and adult weight                  gain are probably a cause of
postmenopausal           breast    cancer.     Body    fatness      is   a     cause     of
postmenopausal           breast     cancer     but    probably       protects      against
premenopausal breast cancer.
Breast tissue comprises mainly fat, glandular tissue (arranged in lobes), ducts, and
connective tissue. Breast tissue develops in response to hormones such as oestrogens,
progesterone, insulin and growth factors. The main periods of development are during
puberty, pregnancy, and lactation. The glandular tissue atrophies after menopause.
Breast cancers are almost all carcinomas of the epithelial cells lining the ducts (the
channels in the breast that carry milk to the nipple).[1] Premenopausal and
postmenopausal breast cancers are considered separately in this Report. Although rare
(Iess than 1 per cent of cases [2]), breast cancer can occur in men, but it is not included
here.
1. Trends, incidence, and survival
Breast cancer is the most common cancer in women in high-, middle- and low-income
countries.[3] Age-adjusted rates of breast cancer in women are increasing in most
countries, particularly in areas where the incidence had previously been low, such as
Japan, China and south-eastern and eastern Europe.[4, 5]
This is predominately a disease of high-income countries where overall rates are nearly
three times higher than in middle- to low-income countries. Around the world, age
adjusted incidence rates range from 75-100 per 100 000 women in North America,
northern Europe, and Australia, to less than 20 per 100 000 in parts of Africa and Asia.
[6] In the USA, rates are higher among white women than those from other ethnic groups,
although mortality is highest in black women.[7]
Overall risk doubles each decade until the menopause, when the increase slows down or
remains stable. However, breast cancer is more corn mon after the menopause. Studies
of women who migrate from areas of low risk to areas of high risk assume the rate in the
host country within one or two generations. This shows that environmental factors are
important in the progression of the disease.[8]
Breast cancers can often be detected at a relatively early stage. In countries that provide
or advocate screening, most of these cancers are diagnosed when the disease is still at a
localised stage.[9] Survival rates range from 90 to less than 50 per cent, depending on
the characteristics of the tumour, its size and spread, and the availability of
treatment.[10] Average 5-year survival rates are more than 80% in North Arnerica,
Sweden, Japan, Finland and Australia compared with Iess than 60 per cent in Brazil and
Slovakia and less than 40 per cent in Algeria.[11] The low survival rate in middle- and
low-income countries can be explained mainly by a lack of early detection programmes,
resulting in a high proportion of women presenting with late-stage disease, as well as by a
Iack of adequate diagnosis and treatment facilities. Breast cancer accounts for nearly 23
per cent of all cancer incidence in wornen and 16 per cent of all cancer deaths (all sites
except for skin (non-melanoma) and in women only). [3, 6] Breast cancer is the ninth
most common cause of death in high income countries and around 69% of all breast
cancer deaths occur in middle- and low-income countries.[3] Mortality rates have
 rernained fairly stable between 1960 and 1990 in most of Europe and the Americas; and
                                                                                           5
</pre>

====================================================================== Einde pagina 103 =================================================================

<br><br>====================================================================== Pagina 104 ======================================================================

<pre> have since showed a decline, which has reached 25-30% in northern Europe.[12] See
 box 1.
 Box 1 cancer incidence and survival
 The cancer incidence rates and figures given in this Report are those reported by cancer registries, now
 established in many countries. These registries record cases of cancer that have been diagnosed. However,
 many cases of cancer are not identified or recorded: some countries do not have cancer registries; regions
 of some countries have few or no records; records in countries suffering war or other disruption are bound
to be incomplete; and some people with cancer do not consult a physician. Altogether, this means that the
 actual incidence of cancer is higher than the figures given here. The cancer survival rates given in this
 chapter and elsewhere are usually overall global averages. Survival rates are generally higher in high
 income countries and other parts of the world where there are established services for screening and early
 detection of cancer and well established treatment facilities. Survival also is often a function of the stage at
 which a cancer is detected and diagnosed. The symptoms of some internal cancers are often evident only
 at a late stage, which accounts for relatively 10w survival rates. In this context, survival’ means that the
 person with diagnosed cancer has not died 5 years after diagnosis.
 2. Pathogenesis
 Breast tissue, as well as hormones and hormone-receptor status, varies at different
stages of life. It is therefore possible that individual risk factors will have different effects
 at different life stages (see 6. Evidence and iudgements). Early menarche, late
 menopause, not bearing children, and late (over 30) first pregnancy all increase breast
 cancer risk.{8, 13] The age when breasts develop, and menopause, are both influenced
 by nutrition, with overnutrition leading to early puberty and late menopause;
 undernutrition delays puberty and advances menopause (see chapter 6.2 Second Expert
 Report).
 Hormones play an important role in breast cancer progression because they modulate
the structure and growth of epithelial tumour cells.[10] Different cancers vary in hormone
sensitivity. Many breast cancers also produce hormones, such as growth factors, that act
 locally, and these can both stimulate and inhibit the tumour’s growth.[14, 15]
 Family history of breast cancer is associated with a 2-3 fold higher risk of the disease.
Some mutations, particularly in BRCA1, BRAC2 and p53 result in a very high risk of breast
cancer. These mutations are rare and account for only 2 to 5 per cent of total cases.{16]
 In addition, growth factor receptor genes, as well as some oncogenes, are overexpressed
 in many breast cancers.[10] (Also see box 2.2. chapter 2, Second Expert Report).
3. Other established causes
3.1 General
This section Iists factors outside the scope of this Report, identified as established
causes of cancer by the World Health Organization International Agency for Research on
Cancer, and other authoritative bodies. These factors are listed in Chapter 2.4 of the
Second Expert Report: tobacco use; infectious agents; radiation; industrial chemicals;
and some medications. Other diseases may also increase the risk of cancer. In the same
way, life events that modify the risk of cancer               —  causative and protective           —  are also
included.
Established’ effectively means ‘beyond reasonable doubt’ roughly the equivalent of the
                                                                          —
judgement of ‘convincing’ used in this Report. Occasionally, authorative findings that
perhaps faIl short of ‘established’ are also included here.
                                                                                                               6
</pre>

====================================================================== Einde pagina 104 =================================================================

<br><br>====================================================================== Pagina 105 ======================================================================

<pre>Where possible, a note of interactive or multiplicative effects with food, nutrition, and the
other factors covered by this Report is added, as is any indication of scale or relative
importance. The factors here are almost all causative, whereas much of the evidence on
food, nutrition, physical activity, and related factors shows or suggests protection against
cancer.
3.2 Specific
Life events. Lifetime exposure to oestrogen, influenced by early menarche, late natural
menopause, not bearing children, and late (over 30) first pregnancy all increase the risk
of, and may be seen as causes of, breast cancer.[8, 13] The reverse also applies: late
menarche, early menopause, bearing children, and early pregnancy all reduce the risk of,
and may be seen as protective against breast cancer. Age of breast development and
menopause are influenced by nutrition, with high-energy diets promoting earlier puberty
and late menopause, and low-energy diets delaying puberty and advancing menopause.
Radiation. lonising radiation exposure from medical treatment such as X-rays,
particularly during puberty, increases risk, even at low doses.[17]
Medication. Hormone replacement therapy is a cause of breast cancer. The increased
risk appears to disappear a few years after cessation.[18] Oral contraceptives containing
both oestrogen and progesterone cause a small, transient, increased risk of breast
cancer; the increased risk disappears after cessation.[19]
4. Interpretation of the evidence specific to breast cancer
4.1 General
For general considerations that may affect interpretatiori of the evidence, see chapters
3.3 and 3.5, and boxes 3.1, 3.2, 3.6 and 3.7 of the Second Expert Report.
‘Relative risk’ is used in this Report to denote ratio measures of effect, including risk
ratios’, ‘rate ratios’, ‘hazard ratios’, and ‘odds ratios’.
4.2 Specific
Considerations specific to breast cancer include:
Patterns. The preponderance of data from high-income countries is a special issue with
breast cancer. Breast cancer is hormone related, and factors that modify risk have
different effects on cancers diagnosed pre- and postmenopause.
Classification. Because of the importance of menopause as an effect modifier, studies
should stratify for menopause status. Many do not.
Confounding. Hormone replacement therapy is an important possible confounder in
postmenopausal breast cancer. A few studies also reported results separately for
different hormone receptor profiles within cancers. High-quality studies adjust for age,
number of reproductive cycles, age at which children were bom, and the taking of
hormone-based medications.
                                                                                            7
</pre>

====================================================================== Einde pagina 105 =================================================================

<br><br>====================================================================== Pagina 106 ======================================================================

<pre>Effect modification. There is growing evidence that the impact of dietary exposures on
risk of breast cancer may differ according to the particular molecular subtypes of cancer.
5. Methodology
To ensure consistency with evidence collected and analysed for the Second Expert Report
much of the methodology following for the Continuous Update Project remains
unchanged from that used previously. Based upon the experience of conducting the
systematic literature reviews for the Second Expert Report some modifications to the
methodology were made. The literature search was restricted to Medline and included
only randomised controlled trials, cohort and case-control studies. The 2008 Continuous
Update Project Breast Cancer SLR inciuded studies published up to December 2007.
Publications in foreign languages were not included. Due to the large number of cohort
studies, analysis and interpretation of case-control studies was not included in the
Continuous Update Project SLR. Meta-analyses and forest plots of highest versus lowest
categories were prepared for breast cancer incidence. Studies with mortality endpoints
previously included in analyses were removed. Studies reporting mean difference as a
measure of association are not included in the Continuous Update Project SLR as relative
risks estimated from the mean differences are not adjusted for possible confounders,
and thus not comparable to adjusted relative risks from other studies. (For more
information on methodology see the full 2008 Continuous Update Project Breast Cancer
SLR (Second Expert Report).
6. Evidence and judgements
The updated search identified 81 new articles, giving a total of 954 publications for
breast cancer. The following sections inciude evidence from case-control studies
considered as part of the Second Expert Report; however as mentioned in the previous
section the evidence from case-control studies was not included in the 2008 Continuous
Update Project Breast Cancer SLR. Fuller summaries of the experimental and
mechanistic evidence can be found in chapters 4-6 of the Second Expert Report. For
information on the criteria for grading the evidence see box 3.8 of the Second Expert
Report. References to studies added in the Continuous Update Project have been
included in the following sections; for details on references to other studies see Sççnçi
 Expert RepQrt.
6.1 Alcoholic drinks
(Also see sections 3.7.1 Alcoholic drinks and 5.4 Alcohol (as ethanol) of the 2008
Continuous Update Project Breast Cancer SLR)
The Continuous Update Project identified 4 new cohort studies[20-23] that investigated
alcoholic drinks and 2 new cohort studies[24, 25] and 3 recent publications from
previously included cohort studies[26-28] that investigated ethanol intake. For
premenopausal breast cancer a total of 4 cohort studies investigated alcoholic drinks
and 6 cohort studies investigated ethanol intake. The respective numbers for
postmenopausal breast cancer were 9 and 16. For all-age breast cancer a total of 13
cohort studies investigated alcoholic drinks and 11 cohort studies investigated ethanol
intake. Most studies showed increased risk with increased intake. Meta-analysis of cohort
studies for the Second Expert Report showed a 10 per cent increased risk for all-age
breast cancer, a 9 per cent increased risk for premenopausal breast cancer and a 8 per
cent increased risk for postmenopausal breast cancer per 10 g ethanol (Page 167
Second Expert Report). An updated meta-analysis for postmenopausal breast cancer
                                                                                           8
</pre>

====================================================================== Einde pagina 106 =================================================================

<br><br>====================================================================== Pagina 107 ======================================================================

<pre>showed an 8 per cent increased risk per 10 g ethanol (Figure Al 2008 Continuous
Update Project Breast Cancer SLR). The Second Expert Report included 31 case-control
studies that investigated alcoholic drinks and 29 case-control studies that investigated
ethanol intake and all-age breast cancer. Meta-analysis of case-control data showed a 5
per cent increased risk per 5 drinks/week, and a 6 per cent increased risk per 10 g
ethanol/day (Pages 166-167 Second Expert Report). Menopausal status did not
significantly alter the association.
Two pooled analyses also showed statistically significant increased risks of 9 and 7 per
cent per 10 g ethanol/day. The first was based on 6 cohort studies with more than 320
000 participants, followed up for up to 11 years, with more than 4300 breast cancer
cases. The other analysed 53 case-control studies, with more than 58 000 cases and
more than 95 000 controls.[29, 30] A meta-analysis of 3 cohort and 7 case-control
studies assessed the association between alcohol intake and the risk of ER-/PR-defined
breast cancer. [31] The dose-response meta-analysis showed that an increase in alcohol
consumption of 10 g of ethanol per day was associated with statistically significant
increased risks for all ER+ (12 per cent), all ER- (7 per cent), ER+PR+ (11 per cent) and
ER+PR- (15 per cent), but not ER-PR-. A statistically significant heterogeneity of the
results across all ER+ versus ER-PR- was observed.
                                                                                        9
</pre>

====================================================================== Einde pagina 107 =================================================================

<br><br>====================================================================== Pagina 108 ======================================================================

<pre> Reactive metabolites of alcohol, such as acetaldehyde, may be carcinogenic.
Additionally, the effects of alcohol may be mediated through the production of
prostaglandins, lipid peroxidation, and the generation of free radical oxygen
species. Alcohol also acts as a solvent, enhancing penetration of carcinogens into
cells. High consumers of alcohol may have diets deficient in essential nutrients,
making tissues susceptible to carcinogenesis. In addition, most experimental
studies in animals have shown that alcohol intake is associated with increased
breast cancer risk. Alcohol interferes with oestrogen metabolism and action in
multiple ways, influencing hormone levels and oestrogen receptors.
There is an interaction between folate and alcohol affecting breast cancer risk:
increased folate status partially mitigates the risk from increased alcohol
consumption.[32]
 The evidence added for the Continuous Update Project is consistent with that
 from the Second Expert Report. There is ample and generally consistent
  evidence from cohort and case-control studies.
 A dose-response relationship is apparent. There is robust evidence for
  mechanisms operating in humans. The conciusion reached for the Second
  Expert Report remains unchanged. The evidence that alcoholic drinks are a
  cause of premenopausal and postmenopausal breast cancer is convincing. No
 threshold was identified.
6.2 Lactation
(Also see section 1.6.1 Breastfeeding of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 2 new cohort studies[33, 34] that
investigated ever having breastfed as compared with never having breastfed and
3 new cohort studies[20, 21, 33] that investigated the total duration of lactation.
For each of premenopausal and postmenopausal breast cancer a total of 2 cohort
studies investigated ever having breastfed compared to never having breastfed
and 2 cohort studies investigated total duration of lactation. For all-age breast
cancer 3 studies investigated ever having breastfed and 6 studies investigated
total duration of lactation. The Second Expert Report included 37 case-control
studies that investigated ever having breastfed as compared with never having
breastfed and 55 case-control studies that investigated the total duration of
lactation. Most cohort and case-control studies reported decreased risk with ever
having breastfed and with increasing duration of breastfeeding. Previous meta
analyses from the Second Expert Report for case-control data showed a 2 per
cent decreased risk per 5 months of total breastfeeding; and for cohort data
showed a non-significant decreased risk (Page 241 Second Expert Report).
Pooled analysis from 47 epidemiological studies in 30 countries (more than 50
000 controls and nearly 97 000 breast cancer cases) showed a statistically
significant decreased risk of breast cancer of 4.3 per cent for each 12 months of
breastfeeding. Menopause status was not an effect modifier.[35] The relationship
between breastfeeding and breast cancer according to hormone receptor status
was investigated in a meta-analysis of 5 population-based case-control studies. A
statistically significantly lower risk was found, both of ER+/PR+ breast cancers
</pre>

====================================================================== Einde pagina 108 =================================================================

<br><br>====================================================================== Pagina 109 ======================================================================

<pre>(22 per cent) and for ER-/PR- cancers (26 per cent), for more than 6 months of
breastfeeding compared with never breastfeeding. [36]
Lactation is associated with increased differentiation of breast cells and with
lower exposure to endogenous sex hormones during amenorrhea accompanying
lactation. In addition, the strong exfoliation of breast tissue during lactation, and
the massive epithelial apoptosis at the end of lactation, could decrease risk by
elimination of cells with potential DNA damage.
  The evidence added for the Continuous Update Project is consistent with that
  from the Second Expert Report. There is abundant epidemiological evidence
  from both cohort and case-control studies, which is consistent and shows a
  dose-response relationship. There is robust evidence for plausible mechanisms
  that operate in humans. The conciusion reached for the Second Expert Report
  remains unchanged. The evidence that lactation protects against both
  premenopausal and postmenopausal breast cancer is convincing.
6.3 Physical activity
(Also see section 6. Physical Activity of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 2 new cohort studies[37, 38]
investigating total physical activity; 1 new cohort study investigating occupational
activity[37]; 3 new cohort studies[37-39] and 1 recent publication from a
previously included cohort study[40j investigating recreational activity; and 2 new
cohort studies[37, 381 investigating household activity. For premenopausal breast
cancer a total of 5 cohort studies investigated total physical activity and 4, 3 and
 1 studies investigated occupational, recreational and household activities
 respectively. For postmenopausal breast cancer 2 studies investigated total
activity and 5, 11 and 1 studies investigated occupational, recreational and
 household activities respectively. For all-age breast cancer 4 studies investigated
total physical activity and 4, 5 and 1 studies investigated occupational,
 recreational and household activities respectively. The Second Expert Report
 included 8 case-control studies that investigated total physical activity, 7 case
 control studies that investigated occupational activity and 11 case-control studies
that i nvestigated recreational activity.
 Menopause age unspecified
 Mast studies showed decreased risk with increased physical activity. Meta
 analysis of case-control studies for the Second Expert Report showed a 10 per
 cent decreased risk per 7 MET-hours recreational activity/ week (Page 204
 Second Expert Report).
 P rem en opa u s e
 Data were inconsistent for cohort studies for physical activity; however mast case
 control studies reviewed for the Second Expert Report showed evidence of
 decreased risk (Page 204 Second Expert Report).
 Postmenopause
 Nearly all of the cohort studies showed decreased risk with increased physical
 activity. The meta-analyses from the Second Expert Report of cohort and case
                                                                                    11
</pre>

====================================================================== Einde pagina 109 =================================================================

<br><br>====================================================================== Pagina 110 ======================================================================

<pre>control data both showed a 3 per cent decreased risk per 7 MET-hours
recreational activity/week (Page 205 Second Expert Report).
Sustained moderate physical activity raises the metabolic rate and increases
maximal oxygen uptake. In the long term, regular periods of such activity increase
the body’s metabolic efficiency and capacity (the amount of work that it can
perform), as well as reducing blood pressure and insulin resistance. In addition, it
decreases levels of oestrogens and androgens in postmenopausal women. Some
trials have also shown decreases in circulating oestrogens, increased menstrual
cycle Iength, and decreased ovulation in premenopausal women with a high level
of physical activity.
  Premenopause: There is ample evidence from prospective studies, but It is
  inconsistent. There is evidence from case-control studies suggestive of a
  decreased risk with higher levels of physical activity. The conclusion reached for
 the Second Expert Report remains unchanged. There is limited evidence
 suggesting that physical activity protects against premenopausal breast cancer.
  Postmenopause: The evidence added in the Continuous Update Project is
  consistent with that from the Second Expert Report. There is ample evidence
 from prospective studies showing lower risk of postmenopausal breast cancer
 with higher levels of physical activity, with a dose-response relationship, although
 there is some heterogeneity. There is little evidence on frequency, duration, or
  intensity of activity. The conclusion reached for the Second Expert Report
  remains unchanged. There is robust evidence for mechanisms operating in
  humans. Physical activity probably protects against postmenopausal breast
  cancer.
6.4 Body fatness
(Also see section 8.1.1 Body Mass Index of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 10 new[34, 41-49] and 2 recent
publications from previously included studies[39, 50] investigating body fatness,
as measured by BMI for pre- and postmenopausal breast cancer. For
premenopausal breast cancer there was a total of 22 studies and for
postmenopausal breast cancer there were 28 studies. The Second Expert Report
included more than 100 case-control studies that investigated body fatness.
When grouped for all ages the Second Expert Report showed that the data were
inconsistent in relationship to body fatness (Page 218 Second Expert Report) and
this remained true for the Continuous Update Project. However, a consistent
effect emerged when they were stratified according to menopausal status.
P rem en opa u s e
Most studies showed a decreased risk for premenopausal breast cancer. Meta
analyses for the Second Expert Report (Page 221 Second Expert Report) showed
a 15 per cent decreased risk per 5kg/m2 for cohort studies and an 8 per cent
decreased risk per 5kg/m
                      2 for case-control studies; the updated meta-analysis for
                                                                                    12
</pre>

====================================================================== Einde pagina 110 =================================================================

<br><br>====================================================================== Pagina 111 ======================================================================

<pre> cohort studies showed a 7 per cent decreased risk per 5kg/m
                                                          2 (Figure BMI4 2008
 Continuous Update Project Breast Cancer SLR). A pooled analysis of four cohort
studies with 723 cases of premenopausal breast cancer followed up for up to 11
years showed a 14 per cent decreased risk per 5kg7m  .[51] A meta-analysis of
                                                     2
 20 cohort studies reported an 8 per cent decreased risk per 5kg/m
                                                              .{52]
                                                              2
 Postmenopause
 Most studies showed an increased risk for postmenopausal breast cancer with
 increased body fatness. Meta-analysis of cohort studies for the Second Expert
 Report (Page 219 Second Expert Report) showed an 8 per cent increased risk per
2 and a 13 per cent increased risk per 5kg/m
5kg/m                                            ; the updated meta-analysis
                                                 2
of cohort studies showed a 13 per cent increased risk per 5kg/m
                                                              2 (Figure BMI7
2008 Continuous Update Project Breast Cancer SLR). A pooled analysis of seven
cohort studies with 3208 cases of postmenopausal breast cancer followed up for
 up to 11 years showed a 9 per cent increased risk per 5kg/m  .[51] A meta
                                                              2
analysis of 31 cohort studies reported a 12 per cent increased risk per
.[52]
2
5kg/m
 Body fatness directly affects levels of many circulating hormones, such as insulin,
insulin-like growth factors, and oestrogens, creating an environment that
encourages carcinogenesis and discourages apoptosis (programmed ceil death).
It also stimulates the body’s infiammatory response, which may contribute to the
initiation and progression of several cancers (see chapter 2.4.1.3 Second Expert
Report). Adjusting for serum levels of oestradiol diminishes or destroys the
association with BMI, suggesting that hormones are a predominant
mechanism.[53]
There is no single well established mechanism though which body fatness could
prevent premenopausal breast cancer. According to the oestrogen plus
progesterone theory, overweight premenopausal women would be protected
because they would be more frequently anovulatory, and therefore less likely to
be exposed to endogenous progesterone. However, this theory is not well
supported by recent studies, which suggest that natural progesterone could be
protective.[54] Normal levels of natural progesterone are likely to be protective,
and well nourished, or perhaps overnourished women, who may become slightly
overweight in adulthood, may be protected by their natural fertile condition.
Another possible mechanism is that the increased adipose tissue-derived
oestrogen levels in overweight children could induce early breast differentiation
and eliminate some targets for malignant transformation.{55] Anovulation and
abnormal hormone profiles are commonly associated with obesity.[56] The age
specific pattern of association of breast cancer with BMI, therefore, is largely
explained by its relationship with endogenous sex hormone levels.
Breast cancer diagnosed after the menopause is much more common. Therefore,
throughout life, a decreased risk of premenopausal breast cancer would be
outweighed by an increased risk of postmenopausal breast cancer.
                                                                                  13
</pre>

====================================================================== Einde pagina 111 =================================================================

<br><br>====================================================================== Pagina 112 ======================================================================

<pre>  Premenopause: The evidence added in the Continuous Update Project is
  consistent with that from the Second Expert Report. There is a substantial
  amount of consistent evidence epidemiological evidence with a dose-response
  relationship, but the mechanistic evidence is speculative. The conciusion
  reached for the Second Expert Report remains unchanged. Greater body fatness
  probably protects against premenopausal breast cancer.
  Postmenopause: The evidence added in the Continuous Update Project is
  consistent with that from the Second Expert Report. There is abundant and
  consistent epidemiological evidence and a dear dose-response relationship with
  robust evidence for mechanisms operating in humans. The conciusion reached
  for the Second Expert Report remains unchanged. The evidence that greater
  body fatness is a cause of postmenopausal breast cancer is convincing.
6.5 Adult attained height
(Also see section 8.3.1 Height of the 2008 Continuous Update Project Breast
Cancer SLR)
The Continuous Update Project identified 5 new cohort studies[34, 39, 41, 48,
57] that investigated adult attained height. The total number of cohort studies
was 21 for all-age or age unspecified, 17 for premenopausal and 22 for
postmenopausal breast cancer. The Second Expert Report inciuded 29 oase-
control studies that investigated adult attained height and all-age breast cancer,
38 for premenopausal and 34 for postmenopausal breast cancer.
Menopausal age unspecified
Most of the studies showed increased risk. Meta-analysis for the Second Expert
Report showed a 9 per cent increased risk per 5cm of height for cohort studies
and a 3 per cent increased risk per 5cm of height for oase-control studies (Page
233 Second Expert Report).
P rem en opa u s e
 Most of the studies showed increased risk. Meta-analysis for the Second Expert
 Report showed a 9 per cent increased risk per 5cm of height for cohort studies
and a 4 per cent increased risk per 5cm for oase-control studies (Page 235
Second Expert Report). An updated meta-analysis of cohort studies also showed a
9 per increased risk per 5cm of height (Figure Htl 2008 Continuous Update
 Project Breast Cancer SLR). A pooled analysis of four cohort studies with 723
 cases of premenopausal breast cancer followed up for up to 11 years showed a
 non-significant increased risk with greater adult attained height.[51]
 Postmenopause
 Nearly all the cohort studies and most oase-control studies showed increased
 risk, with no studies showing statistically significant contrary resuits. Meta
 analyses for the Second Expert Report showed an 11 per cent increased risk per
 5cm of height for cohort studies and a 2 per cent increased risk per 5cm for oase-
 control studies (Page 234 Second Expert Report). An updated meta-analysis
 showed a 10 per increased risk per 5cm of height (Figure Ht4 2008 Continuous
 Update Project Breast Cancer SLR. A pooled analysis of seven cohort studies with
                                                                                 14
</pre>

====================================================================== Einde pagina 112 =================================================================

<br><br>====================================================================== Pagina 113 ======================================================================

<pre>3208 cases of postmenopausal breast cancer followed up for up to 11 years
showed a significantly significant 7 per cent increased risk per 5cm of height.[51]
The general mechanisms through which the factors that lead to greater adult
attained height, or its consequences, could plausibly influence cancer risk are
outlined in chapter 6.21.3 and box 2.4 of the Second Expert Report. Many of
these, such as early-life nutrition, altered hormone profiles, and the rate of sexual
maturation, could plausibly increase cancer risk.
  Premenopause: There are fewer data for premenopausal than for
  postmenopausal breast cancer. The evidence added in the Continuous Update
  Project is consistent with that from the Second Expert Report. The
 epidemiological evidence is generally consistent, with a dose-response
  relationship and evidence for plausible mechanisms. The conclusion reached for
 the Second Expert Report remains unchanged. The factors that lead to greater
 adult height, or its consequences, are probably a cause of premenopausal
  breast cancer. The causal factor is unlikély to be tallness itself, but factors that
  promote linear growth in childhood.
  Postmenopause: The evidence added in the Continuous Update Project is
 consistent with that from the Second Expert Report. There is abundant
 epidemiological evidence, which is generally consistent, with a dear dose
  response relationship and evidence for plausible mechanisms operating in
  humans. The conciusion reached for the Second Expert Report remains
  unchanged. The evidence that the factors that lead to greater adult attained
  height, or its consequences, are a cause of postmenopausal breast cancer is
 convincing. The causal factor is unlikely to be tallness itself, but factors that
  promote linear growth in childhood.
6.6 Abdomina fatness (postmenopause)
(Also see sections 8.2.1 Waist Circumference and 8.2.3. and Waist to hip ratio of
the 2008 Continuous Update Project Breast Cancer SLR)
The Continuous Update Project identified 3 new cohort studies[42, 47, 48] and 1
recent publication from a previously included cohort study[58] that investigated
waist circumference and 3 cohort studies[42, 47, 48] and 2 recent publications
from previously included cohort studies[28, 59] that investigated waist to hip
ratio. In total 9 cohort studies investigated waist circumference and 13 waist to
hip ratio. The Second Expert Report included 3 case-control studies that
investigated waist circumference and 8 that investigated waist to hip ratio.
All of the waist circumference studies and most of those on waist to hip ratio
showed increased risk with increased measures of abdominal fatness. Meta
analysis of cohort studies for the Second Expert Report showed a 5 per cent
increased risk per 8 cm in waist circumference (Page 226 Second Expert Report).
The updated meta-analyses were stratified by whether the study adjusted for BMI.
Studies that did not adjust for BMI showed a 7 per cent increased risk per 8cm in
waist circumference and those that did showed a 4 per cent increased risk
(Figures W5 and W6 2008 Continuous Update Project Breast Cancer SLR).
                                                                                    15
</pre>

====================================================================== Einde pagina 113 =================================================================

<br><br>====================================================================== Pagina 114 ======================================================================

<pre>Meta-analysis of cohort studies for the Second Expert Report showed a 19 per
cent increased risk per 0.1 increment in waist to hip ratio (Page 226 Second
Expert Report). The updated meta-analyses were stratified by whether the study
adjusted for BMI. Studies that did not adjust for BMI showed a 9 per cent
increased risk per 0.1 ncrement in waist to hip ratio and those that did showed a
non-significant increased risk (Figures WHR6 and WHR7 2008 Continuous Update
Project Breast Cancer SLR).
The general mechanisms through which abdominal fatness could plausibly cause
cancer are outlined in chapter 6.1.3 9 and box 2.4 of the Second Expert Report.
The hormonal and other biological effects of being overweight or obese are
outlined in chapter 8 of the Second Expert Report. Many of these, such as
increased levels of circulating oestrogens and decreased insulin sensitivity, are
associated with abdominal fatness independently of overaW body fatness.
 The evidence added in the Continuous Update Project is consistent with that
 from the Second Expert Report. There is a substantial amount of epidemiological
 evidence but some inconsistency. There is robust evidence for mechanisms that
 operate in humans. The conciusion reached for the Second Expert Report
 remains unchanged. Abdominal fatness is a probable cause of postmenopausal
 breast cancer.
6.7 Adult weight gain (postmenopause)
(Also see section 8.1.6 Weight Change of the 2008 Continuous Update Project
Breast Cancer SLR)
The Continuous Update Project identified 3 new cohort studies[42, 47, 48] and 1
recent publication from a previously included cohort study[60] that investigated
adult weight change and postmenopausal breast cancer. In total 10 cohort
studies investigated adult weight change. The Second Expert Report inciuded 17
case-control studies that investigated adult weight change. Nearly all the studies
showed increased risk with increased weight gain in adulthood. Meta-analyses for
the Second Expert Report showed a 3 per cent increased risk per 5kg gained for
the cohort studiesand a 5 per cent increased risk per 5kg for case-control studies
(Page 227 Second Expert Report). Heterogeneity may be explained by failure to
separate postmenopausal women taking hormone replacement therapy.
Body fatness directly affects levels of many circulating hormones, such as insulin,
insulin-like growth factors, and oestrogens, creating an environment that
encourages carcinogenesis and discourages apoptosis (see chapter 2.7.1.3
Second Expert Report). It also stimulates the body’s inflammatory response,
which may contribute to the initiation and progression of several cancers.
 The evidence added in the Continuous Update Project is consistent with that
 from the Second Expert Report. There is ample, consistent epidemiological
 evidence and a dose-response relationship was apparent. The conclusion
 reached for the Second Expert Report remains unchanged. Adult weight gain is a
 probable cause of postmenopausal breast cancer.
                                                                                 16
</pre>

====================================================================== Einde pagina 114 =================================================================

<br><br>====================================================================== Pagina 115 ======================================================================

<pre>6.8 Greater birth weight (premenopause)
(Also see section 8.4.1 Birthweight of the 2008 Continuous Update Project Breast
Cancer SLR)
The Continuous Update Project identified 1 new cohort study[61] that
investigated birth weight and premenopausal breast cancer. In total 6 cohort and
4 case-control studies investigated birth weight. All cohort studies and mast case
control studies showed increased risk with greater birth weight. Meta-analysis of
cohort studies for the Second Expert Report showed an 8 per cent increased risk
per kg (Page 238 Second Expert Report).
The general mechanisms through which the factors that lead to greater birth
weight, or its consequences, could plausibly influence cancer risk are outline in
chapter 6.2.11. of the Second Expert Report many of these, such as long-term
programming of hormonal systems, could plausibly increase cancer risk. Greater
birth weight raises circulating maternal oestrogen levels and may increase insulin
like growth factor (IGF)-1 activity; low birth weight raises fetal and maternal levels
of IGF-1 binding protein. The action of both oestrogens and IGF-1 are thought to
be important in fetal growth and mammary gland development, and play a
central, synergistic role in the initiation and promotion of breast cancer.[62]
Animal experiments also provide evidence that exposure to oestrogens during
fetal and early postnatal development can increase the risk of mammary
cancers.[63]
  The evidence added in the Continuous Update Project is consistent with that
  from the Second Expert Report. There is general consistency amongst the
  relatively few epidemiological studies, with some evidence for a dose-response
  relationship. The mechanistic evidence is speculative. The conciusion reached
  for the Second Expert Report remains unchanged. The factors that lead to
  greater birth weight, or its consequences, are probably a cause of
  premenopausal breast cancer.
6.9 Total fat (postmenopause)
(Also see section 5.2 Total Fat of the 2008 Continuous Update Project Breast
Cancer SLR)
The Continuous Update Project identified 1 new cohort study[64] and 1 recent
publication from a previously included cohort study[65] that investigated total fat
intake and 1 new cohort study[66] and 1 recent publication from a previously
inciuded cohort study{67] that investigated energy from fat and postmenopausal
breast cancer. In total 9 cohort studies investigated total fat intake and 5 cohort
studies investigated energy from fat and postmenopausal breast cancer. The
Second Expert Report inciuded 16 case-control studies that investigated total fat
intake and postmenopausal breast cancer. For total fat most studies showed
increased risk with increased intake. Meta-analyses for the Second Expert Report
showed a non-significant increased risk for cohort studies and an 11 per cent
increased risk per 20g/day for case-control studies (Page 138 Second Expert
 Report). A pooled analysis of cohort studies of more than 7300 cases of breast
cancer showed an overall non-significant decreased risk with increased fat intake.
 Menopausal status did not significantly alter the result.[68] For energy from fat
                                                                                    17
</pre>

====================================================================== Einde pagina 115 =================================================================

<br><br>====================================================================== Pagina 116 ======================================================================

<pre>most cohort studies reported decreased risk with increasing per cent energy from
fat and one reported a statistically significant increased risk.
The Women’s Health Initiative Dietary Modification Randomised Controlled Trial
with 655 cases of postmenopausal breast cancer reported a relative risk of 0.91
(0.83-1.01) for intervention and comparison group after 8.1 years.[69] Adjusting
for change in body weight had no effect on the relative risk. The trial was
designed to reduce fat intake to 20% and increase servings of vegetables and
fruit to 5 per day and increase servings of grains to at east 6 per day. However for
women with at least 36.8% energy from fat at baseline a decrease was observed
for intervention compared with control (RR- 0.78 (0.64-0.95)).
Higher endogenous oestrogen levels after menopause are a known cause of
breast cancer.[53, 70] Dietary fat may also increase endogenous oestrogen
production.[71]
 The evidence added in the Continuous Update Project is consistent with that
 from the Second Expert Report. Evidence from prospective epidemiological
 studies of different types on the whole shows inconsistent effects, while case
 control studies show a significant positive association. Mechanistic evidence is
 speculative. The conciusion reached for the Second Expert Report remains
  unchanged. Overall, there is limited evidence suggesting that consumption of
 total fat is a cause of postmenopausal breast cancer.
6.10 Other exposures
For pre- and postmenopausal breast cancer, other exposures were evaluated.
However, the data were either of too low quality, too inconsistent, or the number
of studies too few to allow conciusions to be reached. The list of exposures is
shown in the matrices under limited no conclusion. Additional meta-analyses of
                                        —
cohort studies on dietary fibre and highest versus lowest category forest plots for
total, red and processed meat, fish, dietary folate and energy were also
conducted as part of the Continuous Update Project (See 2008 Continuous
Update Project Breast Cancer SLR for details).
There is considerable speculation around a biologically plausible interaction of
soy and soya products with breast cancer development, due to their high
phytoestrogen content. Data on pulses (legumes) were sparse and inconsistent.
A meta-analysis of 3 cohort and 6 case-control studies showed a statistically
significant 25 per cent lower risk of breast cancer at any age for highest versus
lowest intake of soy products. [72]
A meta-analysis of 6 cohort and 12 case-control studies reported a statistically
significant 14 per cent lower risk of breast cancer at any age for highest versus
lowest consumption of soy protein (estimated from intake of soy food and dietary
isoflavones). [73] Another meta-analysis reported a statistically significant 12 per
cent lower risk of breast cancer at any age for highest versus lowest intake of
isoflavones.[74] In a subgroup analysis the association was statistically significant
for Asian populations (29 per cent lower risk) but not for Western populations.
[74] These meta-analyses are limited by the difficulty in the standardisation of
                                                                                   18
</pre>

====================================================================== Einde pagina 116 =================================================================

<br><br>====================================================================== Pagina 117 ======================================================================

<pre>measure of soy intake. The quantity and type of soy consumed varied greatly
across the studies, such that the contrasts in intake levels for the reported risk
estimates differed widely. Although results of these meta-analyses suggest that
soy intake is associated with a modest reduction in breast cancer risk,
heterogeneity across studies limits the ability to interpretthe findings.
7. Corn parison with the Second Expert Report
Overall the evidence from the additional cohort studies identified in the
Continuous Update Project was consistent with those reviewed as part of the
Second Expert Report. Much of the new evidence related to body fatness,
abdominal fatness and weight gain; there were also new studies reporting on
alcohol consumption.
8. Conciusions
Since the new evidence that was found as part of the Continuous Update Project
is consistent with the evidence presented in the Second Expert Report the
conclusions are unchanged.
The Continuous Update Project Panel concludes:
The evidence that lactation protects against breast cancer at all ages thereafter is
convincing. Physical activity probably protects against postmenopausal breast
cancer, and there is limited evidence suggesting that it protects against
premenopausal breast cancer. The evidence that alcoholic drinks are a cause of
breast cancer at all ages is convincing. The evidence that the factors that lead to
greater attained adult height or its consequences are a cause of postmenopausal
breast cancer is convincing; these are probably a cause of premenopausal breast
cancer.
The factors that lead to greater birth weight or its consequences are probably a
cause of breast cancer diagnosed premenopause. Adult weight gain is probably a
cause of postmenopausal breast cancer. The evidence that body fatness is a
cause of postmenopausal breast cancer is convincing, and abdominal body
fatness is probably a cause of this cancer. On the other hand, body fatness
probably protects against breast cancer diagnosed premenopause. There is
limited evidence suggesting that total dietary fat is a cause of postmenopausal
breast cancer.
                                                                                 19
</pre>

====================================================================== Einde pagina 117 =================================================================

<br><br>====================================================================== Pagina 118 ======================================================================

<pre>A c k ii 0W 1 e d ge m e n ts
Continuous Update Project Panel Members
          Elisa V Bandera MD PhD
         The Cancer Institute of New Jersey
          New Brunswick, NJ, USA
          David Hunter MBBS ScD
          Harvard University of Public Health
          Boston, MA, USA
         Alan Jackson CBE MD FRCP
          University of Southampton, UK
         John Milner PhD
          National Cancer Institute
          Rockville MD, USA
          HilaryJ Powers PhD RNutr
          University of Sheffield, UK
         Arthur Schatzkin MD DrPH
          National Cancer Institute
          Rockville, MD, USA
          Ricardo Uauy MD PhD
          Instituto de Nutricion y Technologia de los Alimentos
         Santiago, Chile
         Steven Zeisel MD PhD
          University of North Carolina
          Chapel HilI, NC, USA
Observers
          Elio Riboli MD ScM MPH
          Imperial College London, UK
WCRF Executive
          Marilyn Gentry
          President WCRF Global Network
          Kate Allen
          Director, WCRF International
          Deirdre McGinley-Gieser
          Senior Vice President for Programs, AICR
                                                                20
</pre>

====================================================================== Einde pagina 118 =================================================================

<br><br>====================================================================== Pagina 119 ======================================================================

<pre>Secretariat
      Martin Wiseman EROP FRCPath (chair)
      Medical and Scientific Adviser
      WCRF International
      Rachel Thompson PhD PHNutr
      Science Programme Manager (Nutrition)
      WCRF International
      Susan Higginbotham PhD
      Director for Research, AICR
Imperial College London
      Teresa Norat PhD
      Principal Investigator, Continuous Update Project
      Imperial College London
      Doris Chan
      Research Associate, Conti n uous U pdate Project
      Imperial College London
      Rosa Lau
      Research Associate, Continuous Update Project
      Imperial College London
      Rui Veira
      Data Manager, Continuous Update Project
      Imperial College London
                                                        21
</pre>

====================================================================== Einde pagina 119 =================================================================

<br><br>====================================================================== Pagina 120 ======================================================================

<pre>References
1.   Sainsbury JR, Anderson TJ and Morgan DA. ABC of breast diseases: breast
     cancer. BMJ 2000; 321: 745-50.
2.   Fentiman IS, Fourquet A and Hortobagyi GN. Male breast cancer. Lancet
     2006; 367: 595-604.
3.   World Health Organization. The global burden of disease: 2004 update.
     2008.
4.   Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, 2002. CA Cancer
    J Clin 2005; 55: 74-108.
5.   Parkin DM, Whelan SL, Ferlay J, et al. Cancer Incidence in Five Continents,
     Vol. 1 to VIII. Lyon: IARC 2005. International Agency for Research on
     Cancer. Globocan 2002. httpy/wwwdep.iarc.frj. 2006.
6.   International Agency for Research on Cancer. Globocan 2002. http:www
     dep.iarc.fr/. 2006.
7.  Jemal A, Clegg LX, Ward E, et al. Annual report to the nation on the status
     of cancer, 1975-2001, with a special feature regarding survival. Cancer
     2004; 101: 3-27.
8.   McPherson K, Steel CM and Dixon JM. ABC of breast diseases. Breast
     cancer epidemiology, risk factors, and genetics. BMJ 2000; 321: 624-8.
9.   Ries L, Eisner M, Kosary C, et al. SEER Cancer Statistics Review, 1975-
     2002. Bethesda, MD: National Cancer Institute. 2005.
10.  Kufe D, Pollock R, Weichselbaum R, et al. Holland Frei Cancer Medicine. 6
     ed. Hamilton, Ontario: BC Decker. 2003.
11.  Coleman MP, Quaresma M, Berrino F, et al. Cancer survival in five
     continents: a worldwide population-based study (CONCORD). Lancet Oncol
     2008; 9: 730-56.
12.  Bosetti C, Bertuccio P, Levi F, et al. Cancer mortality in the European
     Union, 1970-2003, with a joinpoint analysis. Ann Oncol 2008; 19: 631-40.
13.  MacMahon B. General Motors Cancer Research Prizewinners Laureates
     Lectures. Charles S. Mott Prize. Reproduction and cancer of the breast.
     Cancer 1993; 71: 3 185-8.
14.  Lippman ME, Dickson RB, Gelmann EP, et al. Growth regulation of human
     breast carcinoma occurs through regulated growth factor secretion. J Celi
     Biochem 1987; 35: 1-16.
                                                                                22
</pre>

====================================================================== Einde pagina 120 =================================================================

<br><br>====================================================================== Pagina 121 ======================================================================

<pre>15. Murray PA, Barrett-Lee P, Travers M, et al. The prognostic significance of
    transforming growth factors in human breast cancer. Bri Cancer 1993;
    67: 1408-12.
16. International Agency for Research on Cancer. World Health Report 2008
    Editors P.Boyle, B Levin. Lyon. 2008.
17. Modan B, Chetrit A, Alfandary E, et al. lncreased risk of breast cancer after
    low dose irradiation. Lancet 1989; 1: 629-31.
18. Post-menopausal oestrogen therapy. IARC Monogr Eval Carcinog Risks
    Hum 72:399-530. 1999.
19. International Agency for Research on Cancer. Hormonal Contraception and
    Post-menopausal Hormonal Therapy. In: IARC Monogr Eval Carcinog Risks
    Hum no 72.
    http://monographs.iarc.fr/ENG/Monographs/vo172/volume72.pdf. 1999.
20. Visvanathan K, Crum RM, Strickland PT, et al. Alcohol dehydrogenase
    genetic polymorphisms, low-to-moderate alcohol consumption, and risk of
    breast cancer. Alcohol Clin Exp Res. 2007; 31: 467-76.
21. WirtaIt E, Mattisson 1, Guliberg B, et al. Fat from different foods show
    diverging relations with breast cancer risk in postmenopausal women. Nutr
    Cancer 2005; 53: 135-43.
22. Vogel U, Christensen J, Nexo BA, et al. Peroxisome profilerator-activated
    receptorgamma2 Pro12Ala, interaction with alcohol intake and NSAID use,
    in relation to risk of breast cancer in a prospective study of Danes.
    Carcinogenesis 2007; 28: 427-34.
23. Nielsen NR and Gronbaek M. Interactions between intakes of alcohol and
    postmenopausal hormones on risk of breast cancer. lnti Cancer 2008;
    122: 1109-13.
24. Zhang SM, Lee lM, Manson JE, et al. Alcohol consumption and breast
    cancer risk in the Women’s Health Study. Am J Epidemiol 2007; 165: 667-
    76.
25. Tjonneland A, Christensen J, Olsen A, et al. Alcohol intake and breast
    cancer risk: the European Prospective Investigation into Cancer and
    Nutrition (EPIC). Cancer Causes Control 2007; 18: 361-73.
26. Ericson U, Sonestedt E, Gullberg B, et al. High folate intake is associated
    with lower breast cancer incidence in postmenopausal women in the
    Malmo Diet and Cancer cohort. Am J Clin Nutr 2007; 86: 434-43.
27. Stolzenberg-Solomon RZ, Chang SC, Leitzmann MF, et al. Folate intake,
    alcohol use, and postmenopausal breast cancer risk in the Prostate, Lung,
    Colorectal, and Ovarian Cancer Screening Trial. Am J Clin Nutr 2006; 83:
    895-904.
                                                                                23
</pre>

====================================================================== Einde pagina 121 =================================================================

<br><br>====================================================================== Pagina 122 ======================================================================

<pre>28. Mellemkjaer L, Bigaard J, Tjonneland A, et al. Body composition and breast
    cancer in postmenopausal women: a Danish prospective cohort study.
    Obesity (Silver.Spring) 2006; 14: 1854-62.
29. Smith-Warner SA, Spiegelman D, Yaun SS, et al. Alcohol and breast cancer
    in women: a pooled analysis of cohort studies. JAMA 1998; 279: 535-40.
30. Hamajima N, Hirose K, Tajima K, et al. Alcohol, tobacco and breast
    cancer— collaborative reanalysis of individual data from 53
    epidemiological studies, including 58,515 women with breast cancer and
    95,067 women without the disease. BrJ Cancer 2002; 87: 1234-45.
31. Suzuki R, Orsini N, Mignone L, et al. Alcohol intake and risk of breast
    cancer defined by estrogen and progesterone receptor status--a meta
    analysis of epidemiological studies. Int J Cancer 2008; 122: 1832-41.
32. Boffetta P and Hashibe M. Alcohol and cancer. The Lancet Oncology 2006;
    7: 149-56.
33. Andrieu N, Goldgar DE, Easton DE, et al. Pregnancies, breast-feeding, and
    breast cancer risk in the International BRCA1/2 Carrier Cohort Study
    (IBCCS). J NatI Cancer lnst 2006; 98: 535-44.
34. lwasaki M, Otani T, Inoue M, et al. Body size and risk for breast cancer in
    relation to estrogen and progesterone receptor status in Japan. Ann
    Epidemiol 2007; 17: 304-12.
35. Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer
    and breastfeeding: collaborative reanalysis of individual data from 47
    epidemiological studies in 30 countries, including 50302 women with
     breast cancer and 96973 women without the disease. Lancet 2002; 360:
     187-95.
36.  Ma H, Bernstein L, Ross RK, et al. Hormone-related risk factors for breast
    cancer in women under age 50 years by estrogen and progesterone
     receptor status: results from a case-control and a case-case comparison.
    Breast Cancer Res 2006; 8: R39.
37.  Lahmann PH, Eriedenreich C, Schuit AJ, et al. Physical activity and breast
     cancer risk: the European Prospective Investigation into Cancer and
     Nutrition. Cancer Epidemiol Biomarkers Prev 2007; 16: 36-42.
38. Tehard B, Eriedenreich CM, Oppert JM, et al. Effect of physical activity on
    women at increased risk of breast cancer: results from the E3N cohort
    study. Cancer Epidemiol Biomarkers Prev 2006; 15: 57-64.
39.  Chang SC, Ziegler RG, Dunn B, et al. Association of energy intake and
     energy balance with postmenopausal breast cancer in the prostate, lung,
     colorectal, and ovarian cancer screening trial. Cancer Epidemiol
     Biomarkers Prev 2006; 15: 334-4 1.
                                                                                24
</pre>

====================================================================== Einde pagina 122 =================================================================

<br><br>====================================================================== Pagina 123 ======================================================================

<pre>40. Bardia A, Hattmann LC, Vachon CM, et al. Recreational physical activity
    and risk of postmenopausal breast cancer based en hormone receptor
    status. Arch Intern Med 2006; 166: 2478-83.
41. Lundqvist E, Kaprio J, Verkasalo PK, et al. Co-twin control and cohort
    analyses of body mass index and height in relation to breast, prostate,
    ovarian, corpus uteri, colon and rectal cancer among Swedish and Finnish
    twins. Inti Cancer 2007; 121: 810-8.
42. Palmer JR, ms-Campbell LL, Boggs DA, et al. A prospective study of body
    size and breast cancer in black women. Cancer Epidemiol Biomarkers Prev
    2007; 16: 1795-802.
43. Reeves GK, Pirie K, Beral V, et al. Cancer incidence and mortality in
    relation to body mass index in the Million Women Study: cohort study. BMJ
    2007; 335: 1134.
44. Reinier KS, Vacek PM and Geller BM. Risk factors for breast carcinoma in
    situ versus invasive breast cancer in a prospective study of pre- and post
    menopausal women. Breast Cancer Res Treat 2007; 103: 343-8.
45. Lukanova A, Bjor 0, Kaaks R, et al. Body mass index and cancer: results
    from the Northern Sweden Health and Disease Cohort. Inti Cancer 2006;
    118: 458-66.
46. Michels KB, Terry KL and Willett WC. Longitudinal study on the role of body
    size in premenopausal breast cancer. Arch Intern Med 2006; 166: 2395-
    402.
47. Ahn J, Schatzkin A, Lacey iv, Jr., et al. Adiposity, adult weight change, and
    postmenopausal breast cancer risk. Arch Intern Med 2007; 167: 2091-
    102.
48. Krebs EE, Taylor BC, Cauley JA, et al. Measures of adiposity and risk of
    breast cancer in older postmenopausal women. i Am Geriatr Soc 2006;
    54: 63-9.
49. Gallicchio L, McSorley MA, Newschaffer Ci, et al. Body mass,
    polymorphisms in obesity-related genes, and the risk of developing breast
    cancer among women with benign breast disease. Cancer Detect Prev
    2007; 31: 95-101.
50. Suzuki R, Rylander-Rudqvist T, Ye W, et al. Body weight and
    postmenopausal breast cancer risk defined by estrogen and progesterone
    receptor status among Swedish women: A prospective cohort study. Int i
    Cancer 2006; 119: 1683-9.
51. van den Brandt PA, Spiegelman D, Yaun SS, et al. Pooled analysis of
    prospective cohort studies on height, weight, and breast cancer risk. Am J
    Epidemiol 2000; 152: 514-27.
                                                                                 25
</pre>

====================================================================== Einde pagina 123 =================================================================

<br><br>====================================================================== Pagina 124 ======================================================================

<pre>52. Renehan AG and al. e. Body-mass index and incidence of cancer: a
    systematic review and meta-analysis of prospective observational studies.
    Lancet 2008; 371: 569-78.
53. Key T, Appleby P, Barnes 1, et al. Endogenous sex hormones and breast
    cancer in postmenopausal women: reanalysis of nine prospective studies.
    J Nat! Cancer Inst 2002; 94: 606-16.
54. Campagnoli C, Abba C, Ambroggio S, et al. Pregnancy, progesterone and
    progestins in relation to breast cancer risk. The Journal of Steroid
    Biochemistry and Molecular Biology 2005; 97: 441-50.
55. Hilakivi-Clarke L, Forsen T, Eriksson JG, et al. Tallness and overweight
    during childhood have opposing effects on breast cancer risk. Bri Cancer
    2001; 85: 1680-4.
56. Pasquali R, Pelusi C, Genghini S, et al. Obesity and reproductive disorders
    in women. Human reproductïon update 2003; 9: 359-72.
57. Baer Hi, Rich-Edwards JW, Colditz GA, et al. Adult height, age at attained
    height, and incidence of breast cancer in premenopausal women. lnti
    Cancer 2006; 119:2231-5.
58. Rinaldi S, Key Ti, Peeters PH, et al. Anthropometric measures, endogenous
    sex steroids and breast cancer risk in postmenopausal women: a study
    within the EPIC cohort. lnti Cancer 2006; 118: 2832-9.
59. Tehard B and Clavel-Chapelon F. Several anthropometric measurements
    and breast cancer risk: results of the E3N cohort study. lnti Obes (Lond)
    2006; 30: 156-63.
60. Eliassen AH, Colditz GA, Rosner B, et al. Adult weight change and risk of
    postmenopausal breast cancer. JAMA 2006; 296: 193-201.
      /
61. Michels KB, Xue F, Terry KL, et al. Longitudinal study of birthweight and the
    incidence of breast cancer in adulthood. Carcinogenesis 2006; 27: 2464
    8.
62. Innes K, Byers T and Schymura M. Birth characteristics and subsequent
    risk for breast cancer in very young women. Am i Epidemiol 2000; 152:
    1121-8.
63. Hilakivi-Clarke L. Mechanisms by which high maternal fat intake during
    pregnancy increases breast cancer risk in female rodent offspring. Breast
    Cancer Res Treat 1997; 46: 199-2 14.
64. Lof M, Sandin S, Lagiou P, et al. Dietary fat and breast cancer risk in the
    Swedish women’s lifestyle and health cohort. Bri Cancer 2007; 97: 1570-
    6.
                                                                                26
</pre>

====================================================================== Einde pagina 124 =================================================================

<br><br>====================================================================== Pagina 125 ======================================================================

<pre>65. Sonestedt E, Guliberg B and Wirfalt E. Both food habit change in the past
    and obesity status may influence the association between dietary factors
    and postmenopausal breast cancer. Public Health Nutr 2007; 10: 769-79.
66. Thiebaut AC, Kipnis V, Chang SC, et al. Dietary fat and postmenopausal
    invasive breast cancer in the National Institutes of Health-AARP Diet and
    Health Study cohort. J Nat! Cancer Inst 2007; 99: 451-62.
67. Kim EH, Willett WC, Colditz GA, et al. Nurses’ Health Study (NHS) Cohort
    1976-l996Dietary fat and risk of postmenopausal breast cancer in a 20-
    year follow-up. Am J Epidemiol 2006; 164: 990-7.
68. Smith-Warner SA, Spiegelman D, Adami HO, et al. Types of dietary fat and
    breast cancer: a pooled analysis of cohort studies. Inti Cancer 2001; 92:
    767-74.
69. Prentice RL, Caan B, Chlebowski RT, et al. Low-fat dietary pattern and risk
    of invasive breast cancer: the Women’s Health Initiative Randomized
    Controlled Dietary Modification Trial. JAMA 2006; 295: 629-42.
70. Kaaks R, Berrino F, Key T, et al. Serum sex steroids in premenopausal
    women and breast cancer risk within the European Prospective
    Investigation into Cancer and Nutrition EPIC). J NatI Cancer Inst 2005; 97:
    755-65.
71. Wu AH, Pike MC and Stram DO. Metaanalysis: dietary fat intake, serum
    estrogen levels, and the risk of breast cancer. J Natl Cancer lnst 1999; 91:
    529-34.
72. Qin L, Xu i, Wang P, et al. Soyfood intake in the prevention of breast
    cancer risk in women: a meta-analysis of observational epidemiological
    studies. J NutrSci Vitaminol (Tokyo) 2006; 52: 428-36.
73. Trock B, Hilakivi L and Clarke R. Meta-analysis of soy intake and breast
    cancer risk. J NatI Cancer lnst 2006; 98: 459-7 1.
74. Wu A, Yu M, Tseng C, et al. Epidemiology of soy exposures and breast
    cancer risk. BrJ Cancer 2008; 98: 9-14.
                                                                               27
</pre>

====================================================================== Einde pagina 125 =================================================================

<br><br>====================================================================== Pagina 126 ======================================================================

<pre>Appendix 1 Criteria for grading evidence
(Taken from Chapter 3 of the Second Expert Report)
This box lists the criteria finally agreed by the    Panel that were necessary to support
the judgements shown in the matrices. The            grades shown here are convincing’,
 probable’, limited suggestive’, limited no
                      —                         —    conciusion’, and ‘substantial effect on
risk unlikely’. In effect, the criteria define these terms.
Convincing
These criteria are for evidence strong enough to support a judgement of a convincing
causal relationship, which justifies goals and recommendations designed to reduce
the incidence of cancer. A convincing relationship should be robust enough to be
highly unlikely to be modified in the foreseeable future as new evidence accumulates.
All of the following were generally required:
     •   Evidence from more than one study type.
     •   Evidence from at least two independent cohort studies.
     •   No substantial unexplained heterogeneity within or between study types or in
         different populations relating to the presence or absence of an association, or
         direction of effect.
     •   Good quality studies to exclude with confidence the possibility that the
         observed association resuits from random or systematic error, including
         confounding, measurement error, and selection bias.
     •   Presence of a plausible biological gradient (dose response’) in the
         association. Such a gradient need not be linear or even in the same direction
         across the different levels of exposure, so long as this can be explained
         plausibly.
     •   Strong and plausible experimental evidence, either from human studies or
         relevant animal models, that typical human exposures can lead to relevant
         cancer outcomes.
Proba bio
These criteria are for evidence strong enough to support a judgement of a probabie
causal relationship, which would generally justify goals and recommendations
designed to reduce the incidence of cancer.
All the following were generally required:
     •   Evidence from at least two independent cohort studies, or at least five case
         control studies.
     •   No substantial unexplained heterogeneity between or within study types in the
         presence or absence of an association, or direction of effect.
     •   Good quality studies to exclude with confidence the possibility that the
         observed association resuits from random or systematic error, inciuding
         confounding, measurement error, and selection bias.
     •   Evidence for biological plausibility.
Limited    —   suggestive
These criteria are for evidence that is too limited to permit a probable or convincing
causal judgement, but where there is evidence suggestive of a direction of effect. The
                                                                                          28
</pre>

====================================================================== Einde pagina 126 =================================================================

<br><br>====================================================================== Pagina 127 ======================================================================

<pre>evidence may have methodological flaws, or be limited in amount, but shows a
generally consistent direction of effect. This almost always does not justify
recommendations designed to reduce the incidence of cancer. Any exceptions to this
require special explicitjustification.
All the following were generally required:
     •   Evidence from at least two independent cohort studies or at east five case
         control studies.
     •   The direction of effect is generally consistent though sorne unexplained
         heterogeneity may be present.
     •   Evidence for biological plausibility.
Limited    —  no conclusion
Evidence is so limited that no firm conclusion can be made. This category represents
an entry level, and is intended to allow any exposure for which there are sufficient
data to warrant Panel consideration, but where insufficient evidence exists to permit
a more definitive grading. This does not necessarily mean a limited quantity of
evidence. A body of evidence for a particular exposure might be graded limited      — no
conclusion’ for a number of reasons. The evidence might be limited by the amount of
evidence in terrns of the number of studies available, by inconsistency of direction of
effect, by poor quality of studies (for example, lack of adjustrnent for known
confounders), or by any corn bination of these factors.
When an exposure is graded ‘limited         —  no conclusion’, this does not necessarily
indicate that the Panel has judged that there is evidence of no relationship. With
further good quality research, any exposure graded in this way might in the future be
shown to increase or decrease the risk of cancer. Where there is sufficient evidence
to give confidence that an exposure is unlikely to have an effect on cancer risk, this
exposure will be judged ‘substantial effect on risk unlikely’.
There are also many exposures for which there is such limited evidence that no
judgement is possible. In these cases, evidence is recorded in the full CUP SLRs on
the Diet and Cancer Report website (www.dietandcancerreport.org). However, such
evidence is usually not inciuded in the summaries.
Substantial effect on risk unlikely
Evidence is strong enough to support a judgement that a particular food, nutrition, or
physical activity exposure is unhikely to have a substantial causal relation to a cancer
outcome. The evidence should be robust enough to be unlikely to be modified in the
foreseeable future as new evidence accurnulates.
All of the following were generally required:
      •  Evidence from more than one study type.
      •  Evidence from at least two independent cohort studies.
      •  Summary estimate of effect close to 1.0 for comparison of high versus 10w
         exposure categories.
      •  No substantial unexplained heterogeneity within or between study types or in
         different populations.
      •  Good quality studies to exclude, with confidence, the possibility that the
         absence of an observed association resuits from random or systematic error,
                                                                                      29
</pre>

====================================================================== Einde pagina 127 =================================================================

<br><br>====================================================================== Pagina 128 ======================================================================

<pre>         inciuding inadequate power, imprecision or error in exposure measurement,
         inadequate range of exposure, confounding, and selection bias.
     •  Absence of a demonstrable biological gradient (‘dose response’).
     •  Absence of strong and plausible experimental evidence, either from human
        studies or relevant animal models, that typical human exposures lead to
         relevant cancer outcomes.
 Factors that might misleadingly imply an absence of effect include imprecision of the
exposure assessment, an insufficient range of exposure in the study population, and
inadequate statistical power. Defects in these and other study design attributes
might lead to a false conclusion of no effect.
The presence of a plausible, relevant biological mechanism does not necessarily rule
out a judgement of ‘substantial effect on risk unlikely’. But the presence of robust
evidence from appropriate animal models or in humans that a specific mechanism
exists, or that typical exposures can lead to cancer outcomes, argues against such a
judgement.
 Because of the uncertainty inherent in conciuding that an exposure has no effect on
risk, the criteria used to judge an exposure ‘substantial effect on risk unlikely’ are
roughly equivalent to the criteria used with at least a probable’ level of confidence.
Conclusions of ‘substantial effect on risk unlikely’ with a lower confidence than this
would not be helpful, and could overlap with judgements of limited      —  suggestive or
‘limited no conclusion’.
         —
Special upgrading factors
These are factors that form part of the assessment of the evidence that, when
present, can upgrade the judgement reached. So an exposure that might be deemed
a ‘limited  —  suggestive’ causal factor in the absence, say, of a biological gradient,
might be upgraded to ‘probable’ in its presence. The application of these factors
(listed below) requires judgement, and the way in which these judgements affect the
final conclusion in the matrix are stated.
     •   Presence of a plausible biological gradient (dose response’) in the
        association. Such a gradient need not be linear or even in the same direction
        across the different levels of exposure, so long as this can be explained
         plausibly.
     •  A particularly large summary effect size (an odds ratio or relative risk of 2.0 or
         more, depending on the unit of exposure) after appropriate control for
        confounders.
     •   Evidence from randomised trials in humans.
     •   Evidence from appropriately controlled experiments demonstrating one or
         more plausible and specific mechanisms actually operating in humans.
     •   Robust and reproducible evidence from experimental studies in appropriate
        animal models showing that typical human exposures can lead to relevant
        cancer outcomes.
                                                                                        30
</pre>

====================================================================== Einde pagina 128 =================================================================

<br><br>====================================================================== Pagina 129 ======================================================================

<pre>Le}
(=)
i]
ft
=
S
(=)
=
”
Cc
K=)
a
tot)
et
(5°)
3
n
AA
CD
oO
etn
jee]
=
D
Let)
M”
o
oO
ts)
=
oO
D
=
NO
(=)
ph
(=>)
(2)
=
3
3
5)
=
we

te ai

World Cancer

Research Fund
Hong Kong

World Cancer American Institute World Cancer Wereld Kanker
Research Fund for Cancer Research Research Fund Onderzoek Fonds
Internationa —— :

CAA USER ATE

www.werf.org www.aicr.org www.werf-uk.org www.werf-nl.org www.werf-hk.org

Continuous
Update Project

www.dietandcancerreport.org C U P |

</pre>

====================================================================== Einde pagina 129 =================================================================

<br><br>====================================================================== Pagina 130 ======================================================================

<pre></pre>

====================================================================== Einde pagina 130 =================================================================

<br><br>====================================================================== Pagina 131 ======================================================================

<pre>                                                                                            823852
Van: Germund Daal
Verzonden: vrijdag 21 augustus 2015 17:38
Aan: GR_RGV2O15
Onderwerp: reactie Richtlijnen goede voeding 2015: Alcoholhoudende dranken
Geachte leden van de Commissie Richtlijnen goede voeding 2015,
Het Wereld Kanker Onderzoek Fonds (WKOF) maakt graag van de gelegenheid gebruik om
een reactie te geven op het concept achtergronddocument ‘Alcoholhoudende dranken’ van
de Gezondheidsraad.
Het Wereld Kanker Onderzoek Fonds maakt deel uit van het wereldwijde WCRF netwerk
(World Cancer Research Fund). Het WCRF netwerk is een wereldwijd
samenwerkingsverband van charitatieve organisaties die zich richten op de preventie
van kanker door middel van een gezonde voeding en leefstijl. Het Wereld Kanker Onderzoek
Fonds zet zich in Nederland sinds 1994 in voor kankerpreventie en heeft in 2007 samen met
het WCRF netwerk een baanbrekend rapport gepubliceerd over het verband tussen voeding,
leefstijl en kanker (Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global
Perspective. Hierna aangeduid als het Second Expert Report). Dit is wereldwijd het meest
uitgebreide rapport ooit over kankerpreventie.
Het Wereld Kanker Onderzoek Fonds/WCRF netwerk streeft ernaar continu over de meest
recente inzichten over het verband tussen voeding, leefstiji en kanker te beschikken. Het
Continuous Update Project (CUP) maakt dit mogelijk. Voor
dit doorlopende onderzoeksproject worden alle bevindingen uit wereldwijd onderzoek naar
de preventie en overleving van kanker door middel van voeding, gewicht
en lichaamsbeweging geanalyseerd. Een onafhankelijk onderzoekspanel beoordeelt
doorlopend de nieuwe bevindingen en trekt conclusies hierover. Jaarlijks verschijnen
er nieuwe CUP rapporten op basis van meta-analyses waarbij prospectieve cohort studies de
voorkeur hebben. Kortom, het CUP project van het WCRF netwerk biedt de meest actuele en
grondige analyse van al het onderzoek naar voeding, leefstijl en kankerpreventie.
Vanuit onze specialisatie op het gebied van voeding, leefstijl en kankerpreventie reageert het
Wereld Kanker Onderzoek Fonds graag op de conceptversie van de Gezondheidsraad over
Alcoholhoudende dranken. In de bijlage vindt u onze reactie. De referenties van
de wetenschappelijke rapporten en de achterliggende meta-analyses zijn tevens bijgevoegd,
evenals hyperlinks voor het downloaden van de rapporten. De bestandsgrootte van de
rapporten is namelijk te groot om ze allemaal per mail te sturen.
Het Wereld Kanker Onderzoek Fonds vertrouwt erop dat haar commentaar kan bijdragen aan
de definitieve tekst voor de Richtlijnen goede voeding 2015. Indien u vragen heeft, laat ons
dat dan graag weten.
Met vriendelijke groet,
Germund Daal
Hoofd Communicatie en Gezondheidsvoorlichting
Wereld Kanker Onderzoek Fonds
Leidseplein 33-2, 1017 PS Amsterdam
</pre>

====================================================================== Einde pagina 131 =================================================================

<br><br>====================================================================== Pagina 132 ======================================================================

<pre>www.wkof.nI
samen kanker voorkomen
</pre>

====================================================================== Einde pagina 132 =================================================================

<br><br>====================================================================== Pagina 133 ======================================================================

<pre>Bijlage 1: Reactie Wereld Kanker Onderzoek Fonds
De reactie van het Wereld Kanker Onderzoek Fonds op het achtergronddocument van
de Gezondheidsraad betreft het verband tussen alcohol en verschillende vormen van
kanker (hoofdstuk 3.5 t/m 3.8). Het Wereld Kanker Onderzoek Fonds verwijst in haar
commentaar naar het Second Expert Report en de CUP rapporten (bijlage 2, 4 en 9). Het
CUP rapport Food, Nutrition, Physical Activity, and the Prevention of Breast Cancer is
gebaseerd op een meta-analyse uit 2008. Het CUP rapport Food, Nutrition, Physical
Activity, and the Prevention of Colorectal Cancer is gebaseerd op een meta-analyse uit
2010 (bijlage 5 en 10).
1. Onderscheid tussen typen drank: De Gezondheidsraad maakt in haar onderzoek
naar alcohol en kanker onderscheid tussen bier, wijn en sterke drank (regel 406-73 5).
Het WCRF netwerk evenals het International Agency for Research on Cancer (IARC) van
de Wereldgezondheidsorganisatie maken echter geen onderscheid tussen verschillende
soorten alcoholhoudende dranken omdat het verband tussen alcoholische dranken en
kanker komt door ethanol, ongeacht welk type drank (1,2)• Het Wereld Kanker Onderzoek
Fonds vraagt zich om deze reden af waarom de Gezondheidsraad ervoor heeft gekozen
het onderzoek naar alcohol en kanker op te splitsen in verschillende soorten drank en
adviseert om te kijken naar het verband tussen ethanol en kanker.
2. Darmkanker: De Gezondheidsraad ziet onvoldoende bewijskracht tussen alcohol uit
sterke drank en het risico op darmkanker (regel 491-492). Het WCRF netwerk, evenals
het IARC (2), ziet echter sterk wetenschappelijk bewijs voor een verband tussen alcohol
en dikke darmkanker, ongeacht welk type alcoholische drank (RR 1.10 [1.06-1.13]) (3)•
Het WCRF netwerk schat dat per jaar 7% van de nieuwe gevallen van dikke darm- en
endeldarmkanker in westerse landen voorkomen kan worden door geen alcohol te
drinken (4) Dit zijn jaarlijks in Nederland naar schatting meer dan 1000 gevallen van
darmkanker. Alcohol wordt sinds 1988 door het IARC erkend als “carcinogeen voor
mensen” (indeling in Groep 1) (5) In 2007 heeft het IARC darmkanker toegevoegd als
kankersoort die causaal gerelateerd is aan alcoholgebruik (6) Helaas ontbreken beide
toonaangevende bronnen het WCRF en het IARC in de bronnenlijst van het document
                              —                       —
van de Gezondheidsraad. Het Wereld Kanker Onderzoek Fonds verzoekt de
Gezondheidsraad dan ook om deze wetenschappelijke informatie toe te voegen aan het
achtergrond document.
3. Borstkanker: De Gezondheidsraad vindt geen eenduidig bewijs voor het verband
tussen alcoholische dranken en borstkanker (regel 520-521, 524-525, 528-529). Uit een
analyse van het WCRF netwerk is echter reeds sterk wetenschappelijk bewijs naar
voren gekomen over het verband tussen alcohol en borstkanker bij vrouwen, zowel
premenopauzaal (RR 1.09 [1.01-1.17] (1) als postmenopauzaal (RR 1.08 [1.05-1.11]) (7)
Tevens heeft het IARC in 2007/2010 de conclusie getrokken dat alcohol een risicofactor
is voor borstkanker (2) Het WCRF netwerk schat dat per jaar 22% van de nieuwe
gevallen van borstkanker in westerse landen voorkomen kan worden door geen alcohol
te drinken (4)• Dit zijn jaarlijks in Nederland ongeveer 3200 gevallen. Het Wereld Kanker
Onderzoek Fonds adviseert de Gezondheidsraad om deze wetenschappelijke informatie
toe te voegen aan het achtergrond document en de conclusie over alcohol en
borstkanker te herzien.
4. Andere kankersoorten gerelateerd aan alcohol: Andere kankersoorten die een
relatie hebben met alcohol zijn niet opgenomen in de top 10 ziekten die de
Gezondheidsraad onder de loep heeft genomen en staan dan ook niet in het
achtergronddocument. Echter, uit analyses van het WCRF netwerk is gebleken dat
alcohol een significante risicofactor is voor mond-, keel- en strottenhoofdkanker (RR
</pre>

====================================================================== Einde pagina 133 =================================================================

<br><br>====================================================================== Pagina 134 ======================================================================

<pre>1.03 [1.02-1.04]) (1), slokdarmkanker (RR 1.04 [1.03-1.05]) (1) en leverkanker (RR 1.04
[1.02-1.06]) (8)• Tevens is in een toonaangevende publicatie van het IARC uit 1988 reeds
geconcludeerd dat alcohol het risico op deze kankersoorten significant verhoogt (5) Het
Wereld Kanker Onderzoek Fonds raadt daarom de Gezondheidsraad aan deze
kankersoorten wel te benoemen in het achtergrond document over alcoholhoudende
dranken.
Conciuderend
Zowel het IARC als het WCRF netwerk zijn op basis van analyses van het bestaande
onderzoek tot de conclusie gekomen dat er 7 soorten kanker causaal gerelateerd zijn
aan alcohol, te weten: dikke darmkanker, borstkanker, mondkanker, keelkanker,
strottenhoofdkanker, slokdarmkanker en leverkanker. Het verband tussen alcohol en
kanker is onafhankelijk van het type drank en betreft ethanol.
Het Wereld Kanker Onderzoek Fonds verzoekt de Gezondheidsraad deze informatie,
gebaseerd op wereldwijd wetenschappelijk onderzoek, mee te nemen in de nieuwe
Richtlijnen goede voeding 2015.
Referenties
1. World Cancer Research Fund / American Institute for Cancer Research. Food,
Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective.
Washington DC: AICR, 2007. Ch 4.8 p 160-171. Beschikbaar via:
h ttp://wkofnI/sites/default/files/Secon d-Expert-Report.pdf
2. International Agency for Research on Cancer. IARC monographs on the evaluation of
carcinogenic risks to humans. Volume 96. Alcoholic beverage consumption and ethyl
carbamate (urethane). Lyon: International Agency for Research on Cancer, 2010.
Beschikbaar via: h ttp://m onographs. iarc.fr/ENG/Mon oqraphs/vo196/index.php
3a. World Cancer Research Fund / American Institute for Cancer Research. Continuous
Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of
Colorectal Cancer. 2011. p 20-22. Beschikbaar via:
http://wkofnl/sites/default/flles/Colorectal-Cancer-201 1 -Reportpdf
3b. World Cancer Research Fund / American lnstitute for Cancer Research. Colorectal
Cancer Systematic Literature Review. 2010. Beschikbaar via
h ttp://www. wcrforg/sites/default/files/SLR colorectal can cer 201 0.pdf
4. World Cancer Research Fund / American Institute for Cancer Research. Policy and
Action for Cancer Prevention. Food, Nutrition, Physical Activity and the Prevention of
Cancer: a Global Perspective. Washington DC: AICR, 2009. Beschikbaar via
h ttp://wkofnl/sites/default/fïles/Policy_Report.pdf
5. International Agency for Research on Cancer. IARC Monographs on the Evaluation of
the carcinogenic risks to humans. Volume 44. Alcohol Drinking. Lyon: International
Agency for Research on Cancer, 1988. Beschikbaar via
h ttp://mon ographs. iarc.fr/ENG/Monographs/vo144/mono44.pdf
6. Baan et al. Carcinogenicity of alcoholic beverages. The Lancet Oncology. Volume 8,
2007. p 292-293. Beschikbaar via
http://www.thelancet.com/pdfs/journals/lanonc/PIIS14 70-2045%2807%29 70099-2.pdf
</pre>

====================================================================== Einde pagina 134 =================================================================

<br><br>====================================================================== Pagina 135 ======================================================================

<pre>7a. World Cancer Research Fund / American Institute for Cancer Research. Continuous
Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of Breast
Cancer. 2010. p 8-10. Beschikbaar via h ttp://wkofni/sites/default/files/Breast-Cancer
201 0-Report.vcif
7h. World Cancer Research Fund / American Institute for Cancer Research. Breast
Cancer Systematic Literature Review. 2008. Beschikbaar via
http://www. wcrf ora/sites/defa ult/files/SLR breast ca ncer.pdf
8. World Cancer Research Fund International/American Institute for Cancer Research.
Continuous Update Project Report: Diet, Nutrition, Physical Activity and Liver Cancer.
2015. p 22-27. Beschikbaar via: http://wkofnl/sites/default/files/Liver-Cancer-2015-
Report.pdf
</pre>

====================================================================== Einde pagina 135 =================================================================

<br><br>====================================================================== Pagina 136 ======================================================================

<pre></pre>

====================================================================== Einde pagina 136 =================================================================

<br><br>====================================================================== Pagina 137 ======================================================================

<pre>Alcoholhoudende dranken
GEZONDHEIDSRAAD                                                    Reactie op commentaren
Reactie van de commissie Richtlijnen goede voeding 2015
op het achtergronddocument over Alcoholhoudende
dranken
De commissie heeft op het achtergronddocument over alcoholhoudende dranken
reacties ontvangen van de Federatie Nederlandse Levensmiddelen Industrie (FNLI),
het Kennisinstituut Bier, de Stichting Verantwoorde Alcoholconsumptie (STIVA), het
Trimbos Instituut, het Nederlandse Instituut voor Alcoholbeleid (STAP), het Wereld
Kanker Onderzoek Fonds (WKOF), dat deel uitmaakt van het World Cancer Research
Fund (WCRF) en het Rijksinstituut voor Volksgezondheid en Milieu (RIVM). De
commissie heeft de inhoudelijke reacties betrokken bij het opstellen van het definitieve
achtergronddocument en over het algemeen de tekstuele suggesties overgenomen.
Twee conclusies zijn vervallen naar aanleiding van de ontvangen commentaren. Het
betreft de conclusies over het verband tussen de consumptie van bier en wijn en het
risico op hart- en vaatziekten.
Op de volgende pagina’s beschrijft de commissie in een tabel alle inhoudelijke
commentaren en wat zij daarmee heeft gedaan.
</pre>

====================================================================== Einde pagina 137 =================================================================

<br><br>====================================================================== Pagina 138 ======================================================================

<pre>                Alcoholhoudende dranken
                GEZONDHEIDSRAAD                                                                                                                             Reactie op commentaren
Tabel Overzicht ontvangen inhoudelijke commentaren op achtergronddocument over alcoholhoudende dranken en reactie van de commissie
Commentatoren      Commentaar                                                                           Reactie commissie
FNLI               Het is ons niet duidelijk geworden wat nu precies sterke drank is. Zijn dat alle     Niet verwerkt
                   dranken met een bepaald minimum percentage aan alcohol (en welk percentage           In de aangehaalde publicaties wordt gerapporteerd over bier, wijn en sterke drank,
                   geldt dan)? Is daarbij gecorrigeerd voor de aanwezigheid van andere                  maar zijn geen definities vermeld. Het klopt dat bij de innamegegevens in het
                   voedingsstoffen (zoals bijvoorbeeld suikers in likeurdranken)? Zijn in alle studies  achtergronddocument (tabel 1) de cijfers voor bier inclusief cider zijn en de cijfers
                   de definities hetzelfde?                                                             over wijn inclusief versterkte wijnen; dit is in de voetnoten vermeld. Verschillen in
                   Het is duidelijk bij de innamecijfers dat cider is meegeteld bij bier. Is de         alcoholgehaltes tussen verschillende soorten ‘bier’ en ‘wijn’ zijn wel verdisconteerd
                   voedingswaarde echter gelijk afgezien van het alcoholgehalte? Is er rekening mee     in de gegevens met betrekking tot alcohol uit drank, maar niet in de gegevens met
                   gehouden dat sommige bieren hogere gehaltes aan alcohol hebben dan andere?           betrekking tot de hoeveelheid drank. Mixdranken zijn in Tabel 1 buiten beschouwing
                   We vragen ons bovendien af in hoeverre het meetellen van versterkte wijnen zoals     gelaten.
                   port en sherry bij ‘wijn’ de resultaten niet zullen vertekenen. Daarbij komt dan nog In het beschikbare cohortonderzoek zijn deze definities vaak niet duidelijk
                   dat onduidelijk is of in de studies dezelfde dranken steeds zijn meegeteld als is    beschreven, omdat veel publicaties primair gericht waren op de totale
                   weergegeven in de tabel met wat in Nederland wordt gedronken.                        alcoholinname. De analyses naar type alcoholhoudende drank zijn minder
                   Als laatste vragen we ons af in hoeverre bepaalde mixdrankjes zijn meegenomen        gedetailleerd toegelicht.
                   in het achtergronddocument. Qua hoeveelheid alcohol per 100 gram bevinden
                   deze zich dichter in de buurt van de wijnen dan de sterke drank hoewel ze vaak
                   met sterke drank worden gemaakt.
Kennisinstituut    Allereerst wil ik aangeven dat het eerdere document: ‘Concept                        Niet verwerkt
Bier               Achtergronddocument Richtlijnen goede voeding 2015 - Alcohol’ over het               De keuze om zowel een achtergronddocument Alcohol als een
                   algemeen overeenkomt met onze interpretatie van de huidige wetenschappelijke         achtergronddocument Alcoholhoudende dranken op te stellen, vloeit voort uit de
                                                                                                                                       2
                   stand van zaken. Het huidige achtergronddocument (Alcoholhoudende dranken)           werkwijze van de commissie. De integratie van deze bevindingen gebeurt niet in de
                   vertoont grote discrepanties met het achtergronddocument Alcohol. Dit is             achtergronddocumenten, maar in het advies.
                   overigens niet verrassend omdat de epidemiologie niet in staat is drankspecifieke
                   verschillen goed uit elkaar te trekken. Redenen hiervoor zijn de verschillen tussen
                   een bier— en een wijndrinker die veelal niet worden meegewogen (een
                   belangrijke factor is dieet) en het feit dat mensen nauwelijks alleen bier of alleen
                   wijn drinken.
                   We willen met onderstaand commentaar een constructieve bijdrage leveren aan
                Pagina 2
</pre>

====================================================================== Einde pagina 138 =================================================================

<br><br>====================================================================== Pagina 139 ======================================================================

<pre>                Alcoholhoudende dranken
                GEZONDHEIDSRAAD                                                                                                                   Reactie op commentaren
Commentatoren      Commentaar                                                                       Reactie commissie
                   het document ‘Concept Achtergronddocument Richtlijnen goede voeding 2015 -
                   Alcoholhoudende dranken’. Echter, gezien de genoemde punten en daarmee de
                   zwakheden in het onderzoek naar drankspecifieke effecten, adviseren wij dat u in
                   overweging neemt om op basis van de huidige wetenschappelijke kennis geen
                   onderscheid te maken in alcoholhoudende dranken en het voedingsadvies te
                   richten op alcoholconsumptie in het algemeen zoals in de vorige editie van dé
                   Richtlijnen goede voeding (2006) en zoals ook gedaan wordt in vele andere
                   voedingsadviezen, zoals bijvoorbeeld die in de Dietary Guidelines for Americans,
                   2010.
                   Commentaar 1 (lnterventieonderzoek):
Kennisinstituut    Ondanks uw bewuste keuze voor de intermediairen (bloeddruk, LDL cholesterol      Niet verwerkt
                                                                                                                                                                                 2
Bier               en BMI), willen wij aangeven dat het in het geval van alcoholconsumptie ook      De aangedragen risicofactoren passen niet in de werkwijze van de commissie.
                                                                           3
                   relevant is om te kijken naar HDL cholesterol verhoging ,c.q. HDL gemedieerde
                                     4                                           5-8
                   cholesterol efflux en ook zijn andere beschermende functies . Daarnaast zijn er
                   nog een aantal andere belangrijke factoren die een causaal verband aannemelijk
                                              3        9
                   maken, zoals fibrinogeen en HbAlc niet geëvalueerd. In al deze onderzoeken
                   wordt geen onderscheid gevonden tussen alcoholhoudende dranken, waarmee
                   dus gesuggereerd wordt dat het om een alcoholeffect gaat. lnterventieonderzoek
                   maakt het zeer aannemelijk dat er een causaal verband is tussen matige
                   alcoholconsumptie (dus geen drankspecifieke effecten) en een lagere incidentie
                   van hart- en vaatziekten, zoals besproken in een systematisch review en meta-
                            3                   10
                   analyse en cohort studies .
Kennisinstituut    Wat betreft effect op lichaamsgewicht is er in 2012 een meta-analyse verschenen  Verwerkt
                                                                                         1                                         1
Bier               van Bendsen en collega’s over de relatie bierconsumptie en obesitas.             De publicatie van Bendsen e.a. is toegevoegd aan het achtergronddocument. Dit
                                                                                                    resulteert in een nieuwe paragraaf met een nieuwe conclusie: Er is te weinig
                                                                                                    onderzoek om een uitspraak te doen over het effect van bier op het
                                                                                                    lichaamsgewicht.
                Pagina 3
</pre>

====================================================================== Einde pagina 139 =================================================================

<br><br>====================================================================== Pagina 140 ======================================================================

<pre>                Alcoholhoudende dranken
                GEZONDHEIDSRAAD                                                                                                                               Reactie op commentaren
Commentatoren      Commentaar                                                                               Reactie commissie
                   Commentaar 2 (Cohortonderzoek):
Kennisinstituut    Paragraaf 3.2.1: In het artikel van Ferrari wordt terecht gewezen op het volgende:       Niet verwerkt in paragraaf 3.2.1
Bier               “In this study beer use displayed more apparent risk patterns than wine                  In paragraaf 3.2 was al beschreven voor welke confounders in de beschikbare
                   consumption, particularly in men. Although we believe that this finding is relevant,     onderzoeken werd geadjusteerd. De tekst van deze paragraaf is niet aangepast.
                   we calI for cautious interpretations of these results, as the lifestyle profile of wine  Een kanttekening over verschillen tussen bierdrinkers en wijndrinkers en het risico
                   and beer drinkers is profoundly different.” Hiervoor verwijzen wij door naar             op restconfounding in de analyses specifiek voor type alcoholhoudende drank is
                   commentaar 3 waarin ingegaan wordt op de eetpatronen van bier-, wijn- en                 toegevoegd aan paragraaf 3.1 ‘Methodologische kanttekeningen bij
                   gedistilleerd drinkers.                                                                  cohortonderzoek’.
Kennisinstituut    Bij paragraaf 3.3.1 (Bier) wordt geconcludeerd dat een verband tussen                    Verwerkt
                                                                                                                          11
Bier               bierconsumptie en het risico op hart- en vaatziekten onwaarschijnlijk is, terwijl in     Costanzo e.a.    includeerden in hun meta-analyse zowel cohortonderzoeken als
                   paragraaf 3.3.2 (Wijn) wel uitgebreid in wordt gegaan op de bevindingen uit het          patiëntcontrole onderzoeken. Ook de figuren die de vorm van het verband
                                              11
                   onderzoek van Costanzo        wat betreft wijn en hart- en vaatziekten. Costanzo en      weergeven zijn op deze combinatie van cohortonderzoeken met patiëntcontrole-
                   collega’s schrijven in hun artikel: “Unfortunately, the very limited data available      onderzoeken gebaseerd. Costanzo e.a. rapporteren een subgroepanalyses
                   about either beer or spirit consumption in relation to cardiovascular or total           specifiek over de bevindingen van de cohortonderzoeken, maar daarin zijn
                   mortality, did not allow us to perform a fully meta-analytic investigation on the latter bevindingen ten aanzien van verschillende uitkomstmaten samengevoegd
                   two beverages.” Met dit gegeven is het ons inziens onredelijk om het verband             (coronaire hartziekten, hart- en vaatziekten en totale sterfte). De bevindingen uit het
                                                                                                                                                      12
                   tussen bierconsumptie en het risico op hart- en vaatziekten als onwaarschijnlijk         onderzoek ten aanzien van totale sterfte     betreffen veruit het grootste aantal cases
                   aan te duiden. Ook omdat de auteurs in de discussie specifiek aangeven: “A               (7.208 sterfgevallen), daarom is deze subgroepanalyse niet bruikbaar voor de
                   previous meta-analysis had shown a clear inverse dose-effect curve against               beschrijving van het verband met coronaire hartziekten (4.389 cases) of hart- en
                   vascular events for wine but not for beer intake. Evidence from the current              vaatziekten (1.145 cases). Een cohortonderzoek betreft volgens Costanzo e.a.
                                                                                                                                                                       13
                   updated and extended meta-analysis confirms the significant reduction of overall         myocard infarct, terwijl de publicatie over beroerte gaat.    Bovendien geven
                   vascular risk associated with wine consumption and shows, apparently for the first       Costanzo e.a. aan dat zij de hoeveelheid drank presenteren, terwijl de
                   time, a similar J-shaped relationship between beer intake and cardiovascular risk.       gepresenteerde blootstellingen voor een deel van de publicaties betrekking heeft op
                   Moreover, the comparison of studies which included a parallel, separate                  de hoeveelheid alcohol in de drank. Vanwege genoemde kanttekeningen laat de
                   evaluation of wine and beer consumption, indicates a similar protecting effect of        commissie deze meta-analyse verder buiten beschouwing.
                   either beverage against cardiovascular risk.”
Kennisinstituut    Bij paragraaf 3.4 (Diabetes Mellitus type 2) wordt ons inziens onterecht                 Niet verwerkt
Bier               geconcludeerd dat bierdrinkende mannen een hoger risico hebben op diabetes               Het bespreken van plausibiliteit op basis van mechanistische overwegingen zoals
                Pagina 4
</pre>

====================================================================== Einde pagina 140 =================================================================

<br><br>====================================================================== Pagina 141 ======================================================================

<pre>                Alcoholhoudende dranken
                GEZONDHEIDSRAAD                                                                                                                           Reactie op commentaren
Commentatoren      Commentaar                                                                             Reactie commissie
                                                                                                                                                                                          2
                   mellitus type 2 dan mannen die geen bier drinken en dat er geen verband is             die met betrekking tot adinopectine valt buiten de werkwijze van de commissie. De
                   gevonden bij vrouwen, en dat er alleen met wijnconsumptie een geringe                  commissie baseert zich op verbanden met het ontstaan van chronische ziekten of
                   risicoverlaging is op diabetes mellitus type 2. Wij worden gesterkt in onze mening     met de sterfte en op effecten op bloeddruk, LDL-cholesterol en lichaamsgewicht.
                   door de overall conclusie van dit onderzoek, waarbij vooral wordt ingegaan op een
                   alcoholeffect en niet drankspecifieke effecten en de discussie waarin wordt
                   aangegeven dat mogelijk leefstijl (zoals dieet) het verschil verklaart tussen de bier-
                   en de wijndrinker (zie ook toelichting bij commentaar 3) [het Kennisinstituut Bier
                                                             14
                   citeert de bevindingen van Beulens e.a.      ten aanzien van alcohol en vervolgt dan
                   met de reflectie van deze auteurs over de verschillen tussen bier- en wijndrinkers:]
                   “The specific risk reduction associated with wine consumption, however, appears
                   to contradict the findings of several mechanistic studies. It was previously shown
                   that the reduced risk of diabetes with moderate alcohol consumption can be
                   explained by increased adiponectin concentrations for 25-30%. However,
                   randomized trials in study populations consuming a variety of alcoholic beverages
                   could not detect a difference in the effects on adiponectin concentrations. This
                   suggests that the underlying biological mechanism is most probably explained by
                   alcohol itself.
                   The specific risk reduction observed with wine could thus be attributed to other       Niet verwerkt in paragraaf 3.4
                   factors associated with wine consumption. Previous studies have shown that wine        In paragraaf 3.4 was al beschreven voor welke confounders in de beschikbare
                   drinkers differ from drinkers of other beverages by consuming a healthier diet and     onderzoeken werd geadjusteerd. De tekst van deze paragraaf is niet aangepast.
                   being less likely to smoke. As men and women may also differ with regard to such       Een kanttekening over verschillen tussen bierdrinkers en wijndrinkers en het risico
                   health-related behaviours, as is seen in the different structure of confounders        op restconfounding in de analyses specifiek voor type alcoholhoudende drank is
                   amongst men and women, this could in part explain the specific association             toegevoegd aan paragraaf 3.1 ‘Methodologische kanttekeningen bij
                   observed for wine consumption and the different effects between men and                cohortonderzoek’.
                   women.”
Kennisinstituut    Paragraaf 3.6 Borstkanker: Het feit dat er in het achtergronddocument ‘Alcohol’        Niet verwerkt
Bier               wel een grote bewijskracht wordt gevonden aangaande alcoholconsumptie en               De commissie beschrijft in een ander achtergronddocument de bevindingen ten
                   risico op borstkanker bij vrouwen, terwijl in het achtergronddocument                  aanzien van alcohol. De integratie van conclusies uit de achtergronddocumenten in
                Pagina 5
</pre>

====================================================================== Einde pagina 141 =================================================================

<br><br>====================================================================== Pagina 142 ======================================================================

<pre>                Alcoholhoudende dranken
                GEZONDHEIDSRAAD                                                                                                                           Reactie op commentaren
Commentatoren      Commentaar                                                                              Reactie commissie
                   Alcoholhoudende dranken voor geen van de alcoholhoudende dranken een                    niet aan de orde in de achtergronddocumenten, maar in het advies.
                   eenduidige uitkomst wordt gevonden, is tegenstrijdig. Smith-Warner en collega’s
                   geven aan: “The specific type of alcoholic beverage did not strongly influence risk
                   estimates.” Tjønneland en collega’s concluderen: “This large European study
                   supports previous findings that recent average alcohol intake, irrespectively of
                   beverage type, increases the risk of breast cancer.” Deze bevindingen sterken
                   nogmaals onze overtuiging dat het gaat om een alcoholeffect en dat daarom een
                   voedingsadvies op basis van alcoholconsumptie en niet gespecificeerd per drank
                   de voorkeur heeft.
                   Commentaar 3 (Bier-, wijn- en gedistilleerddrinkers en hun verschillen)
Kennisinstituut    Een belangrijke reden om geen onderscheid te maken tussen een bier-, wijn- en           Niet verwerkt
Bier               gedistilleerddrinker, is omdat deze nagenoeg niet bestaan. Er wordt nauwelijks          In de analyses betreffende de verbanden met de consumptie van bier, wijn of sterke
                   alleen bier of alleen wijn of alleen gedistilleerd gedronken. Dit blijkt ook uit        drank met het risico op chronische ziekten wordt doorgaans geadjusteerd voor het
                                                        15
                   onderzoek van Sluik en collega’s.       Zij hebben deelnemers aan de VCP 2007-          gebruik van andere typen alcoholhoudende drank.
                   2010 ingedeeld naar drankvoorkeur, waarbij als criterium is gebruikt dat als 70%
                   van de consumptie bestond uit wijn, dan wel bier, dan wel gedistilleerd, men
                   respectievelijk een wijn-, bier-, gedistilleerd drinker is. Als het aantal glazen bier,
                   wijn of gedistilleerd niet optelde tot 70%, dan had men geen voorkeur. Op basis
                   van deze, overigens niet officieel bestaande, definitie waren de drankvoorkeuren
                   als volgt:
                                 Voorkeur voor                             Geen voorkeur      Geen alcohol
                                 Bier      Wijn       Gedistilleerd
                   Man           32%       10%        5%                   33%                 20%
                   Vrouw         5%        26%        6%                   22%                 41%
                   Verder hebben Sluik en collega’s in hetzelfde onderzoek gekeken naar de                 Verwerkt
                   eetpatronen van de op deze manier gedefinieerde bier-, wijn- en gedistilleerd           In paragraaf 3.1 ‘Methodologische kanttekeningen bij cohortonderzoek’ is een alinea
                              16
                   drinkers.’    Mensen met een voorkeur voor bier hadden ongezondere                      toegevoegd over verschillen tussen bierdrinkers en wijndrinkers en het risico op
                   eetgewoonten dan mensen met een voorkeur voor wijn. Hierdoor is het dus niet            restconfounding in de analyses specifiek voor type alcoholhoudende drank.
                Pagina 6
</pre>

====================================================================== Einde pagina 142 =================================================================

<br><br>====================================================================== Pagina 143 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                          Reactie op commentaren
Commentatoren    Commentaar                                                                         Reactie commissie
                 uitgesloten dat dieet een confounder is in de relatie tussen alcoholconsumptie en
                 gezondheid, waarvoor veelal niet wordt gecorrigeerd. Dat zou een reden kunnen
                 zijn dat de zogenoemde bierdrinkers er daarom ‘slechter’ vanaf komen dan de
                 zogenoemde wijndrinkers.
                 Ook uit een systematische review door Sluik en collega’s blijkt dat drankvoorkeur  Niet verwerkt
                 gerelateerd is aan eetgewoonten. Zij concluderen dat als er specifiek naar         De systematische review van Sluik en collega’s kan niet worden aangehaald in het
                 drankvoorkeur gekeken wordt in relatie tot gezondheid, voeding zeker moet          achtergronddocument, omdat deze nog niet is verschenen.
                 worden meegenomen als confounder aangezien onderliggende
                 voedingsvoorkeuren vaak eerder gerelateerd zijn aan gezondheid dan het type
                 drank.
STIVA            Er bestaat een groot aantal discrepanties tussen de conclusies getrokken in de     Niet verwerkt
                 achtergronddocumenten ‘alcohol’ en ‘alcoholhoudende dranken’. Daarom menen         De integratie van bevindingen tot een richtlijn ten aanzien van het alcoholgebruik is
                 wij dat conclusies met betrekking tot dranktypen niet kunnen bijdragen aan een     niet aan de orde in het achtergronddocument. Deze gebeurt in het advies.
                 eventueel advies over de consumptie van specifieke dranktypen.
                 Onduidelijk is dus hoe de verschillen tussen de uitkomsten gerapporteerd in het
                 achtergronddocument alcoholhoudende dranken en de uitkomsten gerapporteerd
                 in het achtergronddocument alcohol moeten worden geïnterpreteerd. Men zou
                 immers een grote overeenkomst verwachten tussen de uitkomsten van de beide
                 documenten. Dit is echter niet het geval; er zijn tegenstrijdigheden en veel
                 informatie ontbreekt. De volgende tegenstrijdigheden vallen op in de conclusies
                 van de beide documenten (zie ook Tabel 1). [STIVA beschrijft hierna puntsgewijs
                 9 strijdigheden en voegt een overzichtstabel van conclusies met grote
                 bewijskracht in de achtergronddocumenten ‘alcohol’ en ‘alcoholhoudende dranken’
                 bij.]
STIVA            Een belangrijke methodologische kanttekening die wordt gemist is de correctie      Niet verwerkt
                 voor verstoring (confounding) in de vergelijking tussen de effecten van bier, wijn In de analyses betreffende de verbanden met de consumptie van bier, wijn of sterke
                 en gedistilleerde dranken. Een van de grote problemen bij het bestuderen van de    drank met het risico op chronische ziekten wordt doorgaans geadjusteerd voor het
                 effecten van de afzonderlijk alcoholhoudende dranken is dat de meeste              gebruik van andere typen alcoholhoudende drank en daarnaast voor potentiële
              Pagina 7
</pre>

====================================================================== Einde pagina 143 =================================================================

<br><br>====================================================================== Pagina 144 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                               Reactie op commentaren
Commentatoren    Commentaar                                                                              Reactie commissie
                 consumenten zowel bier als wijn als gedistilleerd drinken; het komt zelden voor dat     confounders zoals geslacht, BMI, opleidingsniveau en rookgedrag. De rekenkundige
                 één dranktype uitsluitend wordt geconsumeerd. Veelal (ook in de studies vermeld         bewerking houdt dus tevens rekening met variaties in deze potentiële confounders.
                 in dit achtergronddocument) wordt een rekenkundige bewerking uitgevoerd om              In het achtergronddocument is voor de gepoolde analyses steeds beschreven voor
                 toch een effect van één specifieke dranksoort te kunnen afleiden. Een dergelijke        welke potentiële confounders in de onderzoeken was geadjusteerd. Voor de meta-
                 bewerking houdt geen rekening met variaties in drinkpatronen (dagelijks wijn of         analyses is dat niet mogelijk, omdat daarbij wordt uitgegaan van de
                 wekelijks bier / voor of bij de maaltijd drinken) en variaties in andere factoren       analyseresultaten uit de oorspronkelijke publicaties, waarbij de mate van adjustering
                 (geslacht; vrouwen drinken meestal wijn / leeftijd) en heeft dus tekortkomingen.        tussen de publicaties verschilt.
                 Een directe vergelijking van de effecten van de drie dranktypen op de gezondheid        STIVA merkt terecht op dat in de publicaties die in dit achtergronddocument zijn
                 uitsluitend door middel van epidemiologisch onderzoek heeft dus grote                   aangehaald, doorgaans niet is geadjusteerd voor drinkpatronen (binge drinken,
                 methodologische nadelen en is dus niet verantwoord te maken.                            alcoholgebruik tijdens de maaltijd of juist buiten de maaltijden), maar dat is evenmin
                                                                                                         het geval in de meeste publicaties over verbanden tussen het totale alcoholgebruik
                                                                                                         en het risico op chronische ziekten. Deze kanttekening is dus van toepassing op
                                                                                                         beide achtergronddocumenten (Alcohol en Alcoholhoudende dranken).
STIVA            Een tweede belangrijke methodologische kanttekening betreft de correctie van de         Verwerkt
                 overige leefstijlfactoren (met name dieet) bij typische bierconsumenten,                In paragraaf 3.1 ‘Methodologische kanttekeningen bij cohortonderzoek’ is een alinea
                 wijnconsumenten en consumenten van sterke drank. Een beroemd voorbeeld is               toegevoegd over verschillen tussen bierdrinkers en wijndrinkers en het risico op
                                             17
                 de studie door Grønbaek , die een duidelijk gezondheidsvoordeel liet zien voor de       restconfounding in de analyses specifiek voor type alcoholhoudende drank.
                 wijndrinker in vergelijking met de bierdrinker en de gedistilleerddrinker. Deze
                 studie is later opnieuw geanalyseerd met een uitgebreidere correctie voor de
                                                          18
                 voeding van de diverse typen drinkers ; door deze correctie verdwenen de
                 verschillen tussen bier, wijn en gedistilleerd helemaal. De rol van confounding in
                 de relatie tussen dranktype en gezondheidsuitkomst is daarna nog eens door deze
                                    19
                 groep bevestigd.      Het is dus zeer waarschijnlijk dat de wijndrinker een andere
                 leefstijl(met name voeding) heeft dan de bierdrinker, waardoor de uitkomsten
                                                                       20
                 worden verstoord. Overigens noemen Ferrari e.a.          dit probleem ook in hun
                 discussie: ‘Although we believe that this finding is relevant, we calI for cautious
                 interpretations of these results, as the lifestyle profile of wine and beer drinkers is
                 profoundly different.’
              Pagina 8
</pre>

====================================================================== Einde pagina 144 =================================================================

<br><br>====================================================================== Pagina 145 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                           Reactie op commentaren
Commentatoren    Commentaar                                                                            Reactie commissie
STIVA            Andere grote onderzoeken en reviews die geen effect van dranktype laten zien (op
                 totale sterfte en coronaire hartziekten (paragrafen 3.2 en 3.3) worden niet mede
                 overwogen in dit achtergronddocument. Deze onderzoeken zijn toegevoegd aan
                                                          21-25
                 de referentielijst van dit commentaar.
                 
                                      21
                      Mukamal e.a.       concluderen: “Among men, consumption of alcohol at least      Niet verwerkt
                      three to four days per week was inversely associated with the risk of            Deze publicatie is geïncludeerd in de meta-analyse van Costanzo e.a. en wordt dus
                      myocardial infarction. Neither the type of beverage nor the proportion           niet als aanvullend cohortonderzoek toegevoegd.
                      consumed with meals substantially altered this association. Men who
                      increased their alcohol consumption by a moderate amount during follow-up
                      had a decreased risk of myocardial infarction.”
                 
                                  22
                      Rimm e.a.      concluderen in hun meta-analyse: “Although most ecological        Niet verwerkt
                                                                                                                                                22
                      studies support the hypothesis that wine consumption is most beneficial, the     De meta-analyse van Rimm e.a. uit 1996      is gedateerd en wordt aangehaald in de
                                                                                                                                    11
                      methodological problems of these studies limit their usefulness in drawing       publicatie van Costanzo e.a.
                      conclusions. Most of the differences in findings regarding specific drink types
                      are probably due to differences in patterns of drinking specific types of
                      alcoholic drink and to differing associations with other risk factors. Results
                      from observational studies, where individual consumption can be assessed in
                      detail and linked directly to coronary heart disease, provide strong evidence
                      that a substantial proportion of the benefits of wine, beer, or spirits are
                      attributable primarily to the alcohol content rather than to other components of
                      each drink.”
                 
                                 23
                      Cleophas      concludeert uit zijn systematische review: “1. Small doses of      Niet verwerkt
                                                                                                                                                        23
                      alcohol (1-4 drinks a day) are associated with a slightly reduced risk of        De systematische review van Cleophas uit 1999       is eveneens gedateerd; deze
                                                                                                                                                                 11
                      mortality and coronary heart disease (CHD). 2. Small doses (1-4 drinks a day)    wordt niet aangehaald in de publicatie van Costanzo e.a.
                      of wine, beer, and spirits are equally beneficial. 3. Apart from a direct
                      beneficial effect of low doses of alcohol on mortality and CHD, some
                      psychological factors may contribute to its beneficial effect.”
                 
                                                                          24
                      Tolstrup en Gronbaek concluderen in hun review : “Finally, there is some         Niet verwerkt
              Pagina 9
</pre>

====================================================================== Einde pagina 145 =================================================================

<br><br>====================================================================== Pagina 146 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                                  Reactie op commentaren
Commentatoren    Commentaar                                                                               Reactie commissie
                                                                                                                                                      24
                      evidence that wine may have more beneficial effects than beer and distilled         De publicatie van Tolstrup en Gronbaek         is geen systematische review.
                      spirits; however, these results are still controversial and may be confounded
                      by personal characteristics and other lifestyle factors such as diet. The
                      inverse association between alcohol intake and CHD is influenced by age,
                      gender, drinking pattern, and possibly by type of alcohol.”
                 
                                   25
                      Klatsky e.a.    concluderen: “We conclude that (1) drinking ethyl alcohol           Niet verwerkt
                                                                                                                                                                                        11
                      apparently protects against coronary disease, and (2) there may be minor            Deze publicatie verscheen ruim voor de meta-analyse van Costanzo e.a.            en wordt
                      additional benefits associated with drinking both beer and wine, but not            niet als aanvullend cohortonderzoek toegevoegd.
                      especially red wine...etc.”
STIVA            De conclusies in het achtergronddocument ‘alcoholhoudende dranken’ in de                 Verwerkt
                                                                                       11                                11
                 paragrafen 3.2 en 3.3 zijn gebaseerd op een enkele meta-analyse          die een         Costanzo e.a.     includeerden in hun meta-analyse zowel cohortonderzoeken als
                 uitgebreidere versie is van een eerdere meta-analyse door grotendeels dezelfde           patiëntcontrole onderzoeken. Ook de figuren die de vorm van het verband
                                         26
                 groep epidemiologen . Door de uitbreiding van de meta-analyse komen de                   weergeven zijn op deze combinatie van cohortonderzoeken met patiëntcontrole-
                                                                       11
                 auteurs tot een herziene conclusie. Costanzo e.a.        concluderen (zie abstract): “In onderzoeken gebaseerd. Costanzo e.a. rapporteren een subgroepanalyse specifiek
                 previous studies evaluating whether different alcoholic beverages would protect          over de bevindingen van de cohortonderzoeken, maar daarin zijn bevindingen ten
                 against cardiovascular disease, a J-shaped relationship for increasing wine              aanzien van verschillende uitkomstmaten samengevoegd (coronaire hartziekten,
                 consumption and vascular risk was found; however a similar association for beer          hart- en vaatziekten en totale sterfte). De bevindingen uit het onderzoek ten aanzien
                                                                                                                             12
                 or spirits could not be established. An updated meta-analysis on the relationship        van totale sterfte    betreffen veruit het grootste aantal cases (7.208 sterfgevallen),
                 between wine, beer or spirit consumption and vascular events was performed. …            daarom is deze subgroepanalyse niet bruikbaar voor de beschrijving van het
                 … From 16 studies, evidence confirms a J-shaped relationship between wine                verband met coronaire hartziekten (4.389 cases) of hart- en vaatziekten (1.145
                 intake and vascular risk. … … Similarly, from 13 studies a J-shaped relationship         cases). Een cohortonderzoek betreft volgens Costanzo e.a. myocard infarct, terwijl
                                                                                                                                              13
                 was apparent for beer.(..). From 12 studies reporting separate data on wine or           de publicatie over beroerte gaat.      Bovendien geven Costanzo e.a. aan dat zij de
                 beer consumption, two closely overlapping dose-response curves were obtained             hoeveelheid drank presenteren, terwijl de gepresenteerde blootstellingen voor een
                 (maximal protection of 33% at 25 g/day of alcohol). This meta-analysis confirms          deel van de publicaties betrekking heeft op de hoeveelheid alcohol in de drank.
                 the J-shaped association between wine consumption and vascular risk and                  Vanwege genoemde kanttekeningen laat de commissie deze meta-analyse verder
                 provides, for the first time, evidence for a similar relationship between beer and       buiten beschouwing.
                 vascular risk. In the meta analysis of 10 studies on spirit consumption and
                 vascular risk, no J-shaped relationship could be found.”
              Pagina 10
</pre>

====================================================================== Einde pagina 146 =================================================================

<br><br>====================================================================== Pagina 147 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                             Reactie op commentaren
Commentatoren    Commentaar                                                                               Reactie commissie
                 De auteurs melden in de discussie bovendien dat data voor bier- en gedistilleerd-
                 consumptie nog steeds beperkt zijn: “Unfortunately, the very limited data available
                 about either beer or spirit consumption in relation to cardiovascular or total
                 mortality, did not allow us to perform a fully meta-analytic investigation on the latter
                 two beverages.”
                 De conclusie geformuleerd door de auteurs is dus anders dan de conclusie
                 weergegeven in het achtergronddocument (paragraaf 3.3.1). Deze laatste is
                 gebaseerd op een andere analyse, een deelanalyse, uit hetzelfde artikel. Het is
                 vooralsnog onduidelijk waarom het achtergronddocument deze analyse volgt en
                 op basis van deze analyse de relatie tussen bierconsumptie en hart- en
                 vaatziekten risico aanduidt als een onwaarschijnlijk verband en niet de
                 uiteindelijke conclusie van de auteurs volgt.
STIVA            Een derde belangrijke methodologische kanttekening wordt terecht gemaakt op              Niet verwerkt
                                                                                                                                                                                          28-33
                 pagina 7, namelijk dat er kritische opmerkingen zijn gemaakt over de controle            De bevindingen uit de referenties met betrekking tot het sick quitters argument      ,
                                                                                      27
                 groepen (geheelonthouder) in cohortonderzoeken (Fillmore e.a. ) naar de                  hebben betrekking op het totale alcoholgebruik en niet op het gebruik van bier, wijn
                 associatie tussen alcohol en ziekte uitkomsten. Het is echter voor de volledigheid       of sterke drank. Daarom passen ze niet in dit achtergronddocument.
                 goed te vermelden dat cohorten die wel een onderscheid hebben kunnen maken
                 tussen niet-drinkers en ex-drinkers in hun controle groep, geen essentiële
                                                                     28-31
                 verschillen vonden in de beschreven associaties.          Het ‘sick quitters’ argument
                 lijkt dus niet te gelden. Ook wanneer de controle niet uit geheelonthouders bestaat
                                                                                                 32,33
                 maar uit lichte drinkers zijn er verdere dalingen van het risico beschreven.
STIVA            Met betrekking tot het interventieonderzoek, begrijpen we de keuze voor de               Niet verwerkt
                 intermediairen (bloeddruk, LDL cholesterol en BMI) zoals die wordt omschreven in         De aangedragen risicofactoren (HDL-cholesterol, HDL gemedieerde cholesterol
                                                                                                                3-8                    34                                                   2
                 het document ‘werkwijze van de commissie richtlijnen goede voeding 2015’. Wij            efflux , fibrinogeen en HbAlc ) passen niet in de werkwijze van de commissie.
                 betreuren de gekozen benadering echter in het geval van dit specifieke
                 achtergronddocument.
                                                                                           4
                 HDL cholesterol verhoging, c.q. HDL gemedieerde cholesterol efflux en andere
                 HDL functies worden niet meegewogen in het hoofdstuk 2: lnterventieonderzoek.
              Pagina 11
</pre>

====================================================================== Einde pagina 147 =================================================================

<br><br>====================================================================== Pagina 148 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                             Reactie op commentaren
Commentatoren    Commentaar                                                                           Reactie commissie
                 Deze keuze is gemaakt omdat medicijnen en niacine die HDL cholesterol
                 verhogen, niet aantoonbaar bijdragen aan het voorkomen van hartaanvallen. Er
                 zijn echter een beperkt aantal geneesmiddelen getest dat HDL cholesterol
                 verhoogt, c.q. HDL functie verbetert en alcohol (net als lichamelijke activiteit) is
                 een van de weinige nutriënten die niet alleen HDL cholesterol verhoogt maar ook
                                                                  5-8
                 zijn beschermende functies positief beïnvloedt.      HDL wordt in dezelfde mate
                                                           3,7
                 verhoogd door bier, wijn en gedistilleerd ,evenals de meeste andere
                                                                                    3
                 intermediairen zoals gerapporteerd in de meta-analyse van Brien .
                                                                                              34
                 Bovendien wordt door het volgen van de cases geëvalueerd door het lOM           een
                 aantal andere belangrijke factoren die een causaal verband aannemelijk maken,
                 zoals fibrinogeen en HbAlc niet geëvalueerd.
                 Door deze benadering kan de commissie geen conclusie trekken over de effecten
                 van alcoholhoudende dranken op geen enkele intermediair (zelfs niet LDL
                 cholesterol, noch bloeddruk). Interventie onderzoek maakt echter zeer
                 aannemelijk dat er een causaal verband is tussen consumptie van matige
                 hoeveelheden alcoholhoudende dranken en een lagere incidentie van hart- en
                                                                                               3
                 vaatziekten, zoals besproken in een systematisch review en meta-analyse en
                                10
                 cohort studies .
STIVA            In het werkwijze document wordt gesteld dat de commissie zich in beginsel            Niet verwerkt
                 beperkt in haar literatuuronderzoek tot een kritische evaluatie van gepoolde         De multicenter studies van de European Prospective Investigation into Cancer and
                                                                                                                                 20
                 analyses, meta-analyses en systematische reviews die gepubliceerd zijn in peer-      Nutrition van Ferrari e.a.    voldoen aan het criterium van een gepoolde analyse. De
                 reviewed tijdschriften. In gepoolde analyses en meta-analyses worden de              commissie beschouwt de cohorten binnen EPIC als onafhankelijke
                 bevindingen uit meerdere oorspronkelijke onderzoeken met overeenkomstige             cohortonderzoeken. In gepoolde analyses worden de risicoschatters van alle
                 vraagstelling en aanpak gecombineerd tot een nieuwe risicoschatting.                 cohorten op identieke wijze geadjusteerd voor potentiële confounders en daarna
                 Echter de conclusies met betrekking tot totale sterfte zijn gebaseerd op één         samengevoegd. In meta-analyses worden risicoschatters samengevoegd zoals die
                                    20
                 multicenter studie , die wellicht voldoet aan het criterium ‘gepoolde analyse’,      in de geïncludeerde publicaties zijn gerapporteerd. Nadeel van een meta-analyse
                 maar niet aan het criterium ‘bevindingen uit meerdere oorspronkelijke                ten opzichte van een gepoolde analyse is, dat de mate van adjustering voor
                 onderzoeken gecombineerd tot een nieuwe risicoschatting’. Toch wordt de              confounders in een meta-analyse verschilt tussen de geïncludeerde
              Pagina 12
</pre>

====================================================================== Einde pagina 148 =================================================================

<br><br>====================================================================== Pagina 149 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                            Reactie op commentaren
Commentatoren    Commentaar                                                                           Reactie commissie
                 bewijskracht als ‘groot’ omschreven.                                                 risicoschattingen; voordeel van een meta-analyse ten opzichte van een gepoolde
                                                                                                      analyse is, dat het in principe mogelijk is om al het relevante beschikbare onderzoek
                                                                                                      te includeren. Zowel meta-analyses als gepoolde analyses kunnen – afhankelijk van
                                                                                                      het beschikbare onderzoek en de bevindingen - aanleiding geven tot conclusies met
                                                                                                      grote bewijskracht.
STIVA            Het is opvallend dat met betrekking tot Diabetes Mellitus type 2 (paragraaf 3.4), op Niet verwerkt
                                                               14                                                                                     35-39
                 basis van het onderzoek van Beulens e.a.         het achtergronddocument conclusies  De opmerking met betrekking tot adinopectine          betreft een hypothese
                 trekt over de verschillende dranktypen, terwijl de auteurs conclusies trekken over   betreffende een werkingsmechanisme. Het past niet in de werkwijze van de
                                                                                                                                         2
                 ‘moderate alcohol consumption’ en niet over drank specifieke effecten. De auteurs    commissie om hier op in te gaan. De conclusies van de commissie op basis van de
                 merken in hun discussie op: “The specific risk reduction associated with wine        gepoolde analyse van Beulens e.a. (European Prospective Investigation into Cancer
                                                                                                                     14
                 consumption, however, appears to contradict the findings of several mechanistic      and Nutrition)    betreffen een substantieel aantal cohorten en cases en zijn
                                                                                                                                                                2
                 studies. It was previously shown that the reduced risk of diabetes with moderate     geformuleerd conform de werkwijze van de commissie.
                 alcohol consumption can be explained by increased adiponectin concentrations for
                         35
                 25-30% . However, randomized trials in study populations consuming a variety of
                 alcoholic beverages could not detect a difference in the effects on adiponectin
                                   36-39
                 concentrations.         This suggests that the underlying biological mechanism is
                 most probably explained by alcohol itself.
                 The specific risk reduction observed with wine could thus be attributed to other     Verwerkt
                 factors associated with wine consumption. Previous studies have shown that wine      In paragraaf 3.1 ‘Methodologische kanttekeningen bij cohortonderzoek’ is een alinea
                 drinkers differ from drinkers of other beverages by consuming a healthier diet and   toegevoegd over verschillen tussen bierdrinkers en wijndrinkers en het risico op
                                               40
                 being less likely to smoke.      As men and women may also differ with regard to     restconfounding in de analyses specifiek voor type alcoholhoudende drank.
                 such health-related behaviours, as is seen in the different structure of confounders
                 amongst men and women, this could in part explain the specific association
                 observed for wine consumption and the different effects between men and
                 women.”
STIVA            In paragraaf 3.4 wordt herhaaldelijk gerefereerd aan ‘aanvullend onderzoek’ van      Niet verwerkt
                            41                                                                                                                                       2
                 Cullmann      en telkens wordt vermeld dat het onderzoek een te beperkt aantal       De beschrijving is conform de werkwijze van de commissie.
                 cases betreft om daar conclusies op te baseren. Wellicht kan dit onderzoek
              Pagina 13
</pre>

====================================================================== Einde pagina 149 =================================================================

<br><br>====================================================================== Pagina 150 ======================================================================

<pre>                 Alcoholhoudende dranken
                GEZONDHEIDSRAAD                                                                                                                          Reactie op commentaren
Commentatoren       Commentaar                                                                         Reactie commissie
                    worden verwijderd of minder worden benadrukt.
Trimbos Instituut   De vraag of gedistilleerd, bier of wijn van invloed is op de morbiditeit in het    Niet verwerkt
                    algemeen en kanker in het bijzonder lijkt ons moeilijk te beantwoorden op basis    In dit achtergronddocument beschrijft de commissie de specifieke bevindingen ten
                    van epidemiologisch onderzoek. Een standaardglas gedistilleerd, bier of wijn bevat aanzien van bier, wijn en sterke drank. In het achtergronddocument Alcohol zijn de
                    allemaal een zelfde hoeveelheid pure alcohol (10 gram in Nederland) en hoewel      bevindingen ten aanzien van het totale alcoholgebruik beschreven. De integratie
                    verschillend qua concentratie alcohol in de drank, leiden ze tot een zelfde BAC in van bevindingen en het opstellen van een richtlijn is aan de orde in het advies.
                    het lichaam. De stof alcohol en het afbraakproduct van alcohol acetaldehyde
                    worden beiden als carcinogeen aangemerkt. Hoewel consumenten voorkeur
                    kunnen hebben voor een specifieke alcoholhoudende drank (wijn, bier of
                    gedistilleerd), worden in de praktijk diverse alcoholhoudende dranken door elkaar
                    heen gedronken (bijvoorbeeld wijn in combinatie met een aperitief en cognac).
Trimbos Instituut   We zijn verbaasd dat de relatie tussen alcohol en borstkanker als niet eenduidig   Niet verwerkt
                    wordt gekenmerkt. We zijn benieuwd wat u precies bedoelt met niet eenduidig.       Het achtergronddocument ‘Alcoholhoudende dranken’ gaat niet over het verbanden
                    Volgens onze lezing van de literatuur is de relatie tussen borstkanker en          met het totale alcoholgebruik, maar over verbanden met het gebruik van bier, wijn of
                    alcoholconsumptie wel eenduidig en dit is onder andere gebaseerd op                sterke drank. Voor deze specifieke typen alcoholhoudende drank heeft de
                                            42-45
                    bijgevoegde literatuur.                                                            commissie geen eenduidige verbanden met het risico op borstkanker gevonden. De
STAP                Wij hebben ernstige twijfels bij de conclusie in het rapport dat er geen eenduidig commissie is het met u eens dat het verband van een hoger alcoholgebruik met een
                    verband zou bestaan tussen alcoholgebruik en het ontstaan van borstkanker bij      hoger risico op borstkanker grote bewijskracht heeft; dit is beschreven in het
                    vrouwen. In de bijlage bij deze reactie sturen we twee recente artikelen mee       achtergronddocument ‘Alcohol’.
                    waarin duidelijke uitspraken worden gedaan over de samenhang tussen                De integratie van bevindingen en het opstellen van een richtlijn is niet aan de orde in
                    alcoholgebruik en het ontstaan van borstkanker en waarbij geen sprake is van het   de achtergronddocumenten, maar in het advies.
                    ontbreken van een eenduidig verband.                                               De door Trimbos bijgevoegde referenties zijn niet toegevoegd om de volgende
                    Wat ons verontrust is dat in uw rapport steeds nadrukkelijk onderscheid wordt      redenen:
                                                                                                       
                                                                                                                     42
                    gemaakt tussen wat bekend is over de relatie tussen borstkanker (en ook in geval        Cao e.a.    verscheen in juli 2015 en betreft analyses van resultaten van de
                    van andere ziekten) en wijngebruik, borstkanker en biergebruik en borstkanker en        twee afzonderlijke cohorten. Het is geen gepoolde analyse of meta-analyse.
                    het gebruik van sterke drank. En dat terwijl algemeen bekend is dat het primair         Deze publicatie valt buiten de periode van het literatuuronderzoek voor het
                    gaat om de relatie tussen het gebruik van alcohol als carcinogene stof en het           achtergronddocument en wordt daarom niet meegenomen.
                    ontstaan van diverse ziekten waaronder kanker. Het kan niet zo zijn dat we als         De publicatie van Hamajima e.a. (van de Collaborative Group on Hormonal
                 Pagina 14
</pre>

====================================================================== Einde pagina 150 =================================================================

<br><br>====================================================================== Pagina 151 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                                Reactie op commentaren
Commentatoren    Commentaar                                                                            Reactie commissie
                 resultaat van uw rapport gaan zien dat de diverse soorten alcoholhoudende drank            Factors in Breast Cancer) betreft het totale alcoholgebruik en niet het gebruik
                 wat de risico’s van het gebruik ervan betreft, tegen elkaar uitgespeeld worden.            van bier, wijn of sterke drank. Daarom is deze publicatie niet toegevoegd in het
                                                                                                                                     43
                 terwijl de kern is dat de alcohol die deze dranken bevatten als zodanig de                 achtergronddocument.        (Deze publicatie is verwerkt in het
                 belangrijkste factor die bepalend is voor de risico’s. Ik vraag u tenminste een apart      achtergronddocument Alcohol.)
                 hoofdstuk te wijden aan de samenhang tussen de besproken ziekten en het                   De meta-analyse van Key e.a. blijft buiten beschouwing omdat daarin
                                                                                                                                                                                          44
                 gebruik van alcohol zonder daarbij onderscheid te maken tussen de diverse                  patiëntcontroleonderzoeken en cohortonderzoeken zijn samengevoegd.
                                                                                                                                                 45
                 verschijningsvormen van alcohol. Dit mede gezien het feit dat er hoe langer hoe       De WCRF-publicatie over borstkanker          betreft het totale alcoholgebruik en niet het
                 meer tussenproducten geconsumeerd worden, zoals Desperados (bier met                  gebruik van bier, wijn of sterke drank. Daarom is deze publicatie niet toegevoegd in
                 Tequila), en Muscat (wijn met een scheutje gedistilleerde alcohol).                   het achtergronddocument. (Deze publicatie is wel toegevoegd aan het
                                                                                                       achtergronddocument Alcohol.)
                                                                                         45
                 Ik stuur u nogmaals het CUP rapport van het WCRF over borstkanker          en de      Niet verwerkt:
                                                                                                       
                                                                                                                                                      45
                 twee zeer recente artikelen van studies die het verband tussen alcohol en                  De WCRF-publicatie over borstkanker          betreft het totale alcoholgebruik en niet
                                                    42,46
                 borstkanker eenduidig aantonen          .                                                  het gebruik van bier, wijn of sterke drank. Daarom is deze publicatie niet
                                                                                                            toegevoegd in het achtergronddocument. (Deze publicatie is wel toegevoegd
                                                                                                            aan het achtergronddocument Alcohol.)
                                                                                                       
                                                                                                                      42
                                                                                                            Cao e.a.     verscheen in juli 2015 en betreft analyses van resultaten van de
                                                                                                            twee afzonderlijke cohorten. Het is geen gepoolde analyse of meta-analyse.
                                                                                                            Deze publicatie valt buiten de periode van het literatuuronderzoek voor het
                                                                                                            achtergronddocument en wordt daarom niet meegenomen.
                                                                                                       
                                                                                                                          46
                                                                                                            Romieu e.a.      (de 11-jaar follow-up van EPIC) presenteren geen analyses voor
                                                                                                            verbanden van het gebruik van bier, wijn of sterke drank met het risico op
                                                                                                            borstkanker en is daarom niet toegevoegd in het achtergronddocument over
                                                                                                            alcoholhoudende dranken. (Deze publicatie is wel toegevoegd aan het
                                                                                                            achtergronddocument Alcohol en vervangt daar de publicatie van Tjonneland
                                                                                                                 47
                                                                                                            e.a.    over de 6-jaar follow-up van EPIC.)
WKOF                  1.   Onderscheid tussen typen drank:                                             Niet verwerkt
                 De Gezondheidsraad maakt in haar onderzoek naar alcohol en kanker                     De keuze om zowel een achtergronddocument Alcohol als een
                 onderscheid tussen bier, wijn en sterke drank (regel 406-735). Het WCRF netwerk       achtergronddocument Alcoholhoudende dranken op te stellen, vloeit voort uit de
              Pagina 15
</pre>

====================================================================== Einde pagina 151 =================================================================

<br><br>====================================================================== Pagina 152 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                        Reactie op commentaren
Commentatoren    Commentaar                                                                           Reactie commissie
                                                                                                                                  2
                 evenals het International Agency for Research on Cancer (IARC) van de                werkwijze van de commissie. De integratie van deze bevindingen gebeurt niet in de
                 Wereldgezondheidsorganisatie maken echter geen onderscheid tussen                    achtergronddocumenten, maar in het advies.
                 verschillende soorten alcoholhoudende dranken omdat het verband tussen
                                                                                                48,49
                 alcoholische dranken en kanker komt door ethanol, ongeacht welk type drank.
                 Het Wereld Kanker Onderzoek Fonds vraagt zich om deze reden af waarom de
                 Gezondheidsraad ervoor heeft gekozen het onderzoek naar alcohol en kanker op
                 te splitsen in verschillende soorten drank en adviseert om te kijken naar het
                 verband tussen ethanol en kanker.
WKOF                  2.   Darmkanker:                                                                Niet verwerkt
                 De Gezondheidsraad ziet onvoldoende bewijskracht tussen alcohol uit sterke           De keuze om zowel een achtergronddocument Alcohol als een
                 drank en het risico op darmkanker (regel 491-492). Het WCRF netwerk, evenals         achtergronddocument Alcoholhoudende dranken op te stellen, vloeit voort uit de
                           48                                                                                                     2
                 het IARC , ziet echter sterk wetenschappelijk bewijs voor een verband tussen         werkwijze van de commissie. De commissie concludeert in het
                 alcohol en dikke darmkanker, ongeacht welk type alcoholische drank (RR 1.10          achtergronddocument Alcohol dat een alcoholgebruik van 30 tot 60 versus 0 gram
                               50,51
                 [1.06-1.13]).       Het WCRF netwerk schat dat per jaar 7% van de nieuwe gevallen    per dag samenhangt met een ongeveer 20 procent hoger risico op darmkanker.
                 van dikke darm- en endeldarmkanker in westerse landen voorkomen kan worden           De integratie van deze bevindingen gebeurt niet in de achtergronddocumenten,
                                                52
                 door geen alcohol te drinken. Dit zijn jaarlijks in Nederland naar schatting meer    maar in het advies.
                 dan 1000 gevallen van darmkanker. Alcohol wordt sinds 1988 door het IARC
                                                                              53
                 erkend als ‘carcinogeen voor mensen’ (indeling in Groep 1). In 2007 heeft het
                 IARC darmkanker toegevoegd als kankersoort die causaal gerelateerd is aan
                                   54
                 alcoholgebruik. Helaas ontbreken beide toonaangevende bronnen het WCRF en
                                                                                       —
                 het IARC in de bronnenlijst van het document van de Gezondheidsraad. Het
                            —
                 Wereld Kanker Onderzoek Fonds verzoekt de Gezondheidsraad dan ook om deze
                 wetenschappelijke informatie toe te voegen aan het achtergrond document.
WKOF                  3.   Borstkanker:                                                               Niet verwerkt
                 De Gezondheidsraad vindt geen eenduidig bewijs voor het verband tussen               De keuze om zowel een achtergronddocument Alcohol als een
                 alcoholische dranken en borstkanker (regel 520-521, 524-525, 528-529). Uit een       achtergronddocument Alcoholhoudende dranken op te stellen, vloeit voort uit de
                                                                                                                                  2
                 analyse van het WCRF netwerk is echter reeds sterk wetenschappelijk bewijs naar      werkwijze van de commissie. De commissie concludeert in het
                 voren gekomen over het verband tussen alcohol en borstkanker bij vrouwen,            achtergronddocument Alcohol dat een alcoholgebruik vanaf 10 gram versus 0 per
              Pagina 16
</pre>

====================================================================== Einde pagina 152 =================================================================

<br><br>====================================================================== Pagina 153 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                                                                                            Reactie op commentaren
Commentatoren    Commentaar                                                                              Reactie commissie
                                                                  52
                 zowel premenopauzaal (RR 1.09 [1.01-1.17])           als postmenopauzaal (RR 1.08       dag hangt samen met een ongeveer 5% hoger risico op borstkanker bij vrouwen.
                               45,55
                 [1.05-1.11]).       Tevens heeft het IARC in 2007/2010 de conclusie getrokken dat       De integratie van deze bevindingen gebeurt niet in de achtergronddocumenten,
                                                                48
                 alcohol een risicofactor is voor borstkanker.     Het WCRF netwerk schat dat per        maar in het advies.
                 jaar 22% van de nieuwe gevallen van borstkanker in westerse landen voorkomen
                                                             52
                 kan worden door geen alcohol te drinken.       Dit zijn jaarlijks in Nederland ongeveer
                 3200 gevallen. Het Wereld Kanker Onderzoek Fonds adviseert de
                 Gezondheidsraad om deze wetenschappelijke informatie toe te voegen aan het
                 achtergrond document en de conclusie over alcohol en borstkanker te herzien.
WKOF                  4.    Andere kankersoorten gerelateerd aan alcohol:                                Niet verwerkt
                 Andere kankersoorten die een relatie hebben met alcohol zijn niet opgenomen in          Deze informatie over andere kankersoorten dan borstkanker, darmkanker en
                 de top 10 ziekten die de Gezondheidsraad onder de loep heeft genomen en staan           longkanker past niet bij de werkwijze die de commissie volgt bij het opstellen van de
                                                                                                                                   2
                 dan ook niet in het achtergronddocument. Echter, uit analyses van het WCRF              achtergronddocumenten.
                 netwerk is gebleken dat alcohol een significante risicofactor is voor mond-, keel-
                                                                   49
                 en strottenhoofd-kanker (RR 1.03 [1.02-1.04]) , slokdarmkanker (RR 1.04 [1.03-
                        49                                        56
                 1.05])    en leverkanker (RR 1.04 [1.02-1.06]) . Tevens is in een toonaangevende
                 publicatie van het IARC uit 1988 reeds geconcludeerd dat alcohol het risico op
                                                            53
                 deze kankersoorten significant verhoogt.      Het Wereld Kanker Onderzoek Fonds
                 raadt daarom de Gezondheidsraad aan deze kankersoorten wel te benoemen in
                 het achtergrond document over alcoholhoudende dranken.
                 Concluderend
                 Zowel het IARC als het WCRF netwerk zijn op basis van analyses van het
                 bestaande onderzoek tot de conclusie gekomen dat er 7 soorten kanker causaal
                 gerelateerd zijn aan alcohol, te weten: dikke darmkanker, borstkanker,
                 mondkanker, keelkanker, strottenhoofdkanker, slokdarmkanker en leverkanker.
                 Het verband tussen alcohol en kanker is onafhankelijk van het type drank en
                 betreft ethanol.
                 Het Wereld Kanker Onderzoek Fonds verzoekt de Gezondheidsraad deze
                 informatie, gebaseerd op wereldwijd wetenschappelijk onderzoek, mee te nemen
              Pagina 17
</pre>

====================================================================== Einde pagina 153 =================================================================

<br><br>====================================================================== Pagina 154 ======================================================================

<pre>              Alcoholhoudende dranken
             GEZONDHEIDSRAAD                                                    Reactie op commentaren
Commentatoren    Commentaar                                   Reactie commissie
                 in de nieuwe Richtlijnen goede voeding 2015.
              Pagina 18
</pre>

====================================================================== Einde pagina 154 =================================================================

<br><br>====================================================================== Pagina 155 ======================================================================

<pre>Alcoholhoudende dranken
GEZONDHEIDSRAAD                                                                   Reactie op commentaren
Literatuur
1     Bendsen NT, Christensen R, Bartels EM, Kok FJ, Sierksma A, Raben A e.a. Is beer consumption
      related to measures of abdominal and general obesity? A systematic review and meta-analysis.
      Nutr Rev 2013; 71(2): 67-87.
2     Gezondheidsraad. Werkwijze van de commissie Richtlijnen goede voeding 2015 -
      Achtergronddocument bij Richtlijnen goede voeding 2015. Den Haag: Gezondheidsraad; 2015:
      publicatienr. A15/03.
3     Brien SE, Ronksley PE, Turner BJ, Mukamal KJ, Ghali WA. Effect of alcohol consumption on
      biological markers associated with risk of coronary heart disease: systematic review and meta-
      analysis of interventional studies. BMJ 2011; 342: d636.
4     Rohatgi A, Khera A, Berry JD, Givens EG, Ayers CR, Wedin KE e.a. HDL cholesterol efflux
      capacity and incident cardiovascular events. N Engl J Med 2014; 371(25): 2383-2393.
5     Sierksma A, van der Gaag MS, van TA, James RW, Hendriks HF. Kinetics of HDL cholesterol and
      paraoxonase activity in moderate alcohol consumers. Alcohol Clin Exp Res 2002; 26(9): 1430-
      1435.
6     Sierksma A, Vermunt SH, Lankhuizen IM, van der Gaag MS, Scheek LM, Grobbee DE e.a. Effect
      of moderate alcohol consumption on parameters of reverse cholesterol transport in
      postmenopausal women. Alcohol Clin Exp Res 2004; 28(4): 662-666.
7     van der Gaag MS, van TA, Scheek LM, James RW, Urgert R, Schaafsma G e.a. Daily moderate
      alcohol consumption increases serum paraoxonase activity; a diet-controlled, randomised
      intervention study in middle-aged men. Atherosclerosis 1999; 147(2): 405-410.
8     van der Gaag MS, van TA, Vermunt SH, Scheek LM, Schaafsma G, Hendriks HF. Alcohol
      consumption stimulates early steps in reverse cholesterol transport. J Lipid Res 2001; 42(12):
      2077-2083.
9     Schrieks IC, Heil AL, Hendriks HF, Mukamal KJ, Beulens JW. The effect of alcohol consumption on
      insulin sensitivity and glycemic status: a systematic review and meta-analysis of intervention
      studies. Diabetes Care 2015; 38(4): 723-732.
10    Mukamal KJ, Jensen MK, Gronbaek M, Stampfer MJ, Manson JE, Pischon T e.a. Drinking
      frequency, mediating biomarkers, and risk of myocardial infarction in women and men. Circulation
      2005; 112(10): 1406-1413.
Pagina 19
</pre>

====================================================================== Einde pagina 155 =================================================================

<br><br>====================================================================== Pagina 156 ======================================================================

<pre>Alcoholhoudende dranken
GEZONDHEIDSRAAD                                                                    Reactie op commentaren
11    Costanzo S, Di CA, Donati MB, Iacoviello L, de GG. Wine, beer or spirit drinking in relation to fatal
      and non-fatal cardiovascular events: a meta-analysis. Eur J Epidemiol 2011; 26(11): 833-850.
12    Nielsen NR, Schnohr P, Jensen G, Gronbaek M. Is the relationship between type of alcohol and
      mortality influenced by socio-economic status? J Intern Med 2004; 255(2): 280-288.
13    Lu M, Ye W, Adami HO, Weiderpass E. Stroke incidence in women under 60 years of age related
      to alcohol intake and smoking habit. Cerebrovasc Dis 2008; 25(6): 517-525.
14    Beulens JW, van der Schouw YT, Bergmann MM, Rohrmann S, Schulze MB, Buijsse B e.a. Alcohol
      consumption and risk of type 2 diabetes in European men and women: influence of beverage type
      and body size The EPIC-InterAct study. J Intern Med 2012; 272(4): 358-370.
15    Sluik D, van LL, Geelen A, Feskens EJ. Alcoholic beverage preference and diet in a representative
      Dutch population: the Dutch national food consumption survey 2007-2010. Eur J Clin Nutr 2014;
      68(3): 287-294.
16    Sluik D, Bezemer R, Sierksma A, Feskens E. Alcoholic Beverage Preference and Dietary Habits: A
      Systematic Literature Review. Crit Rev Food Sci Nutr 2015; 0.
17    Gronbaek M, Deis A, Sorensen TI, Becker U, Schnohr P, Jensen G. Mortality associated with
      moderate intakes of wine, beer, or spirits. BMJ 1995; 310(6988): 1165-1169.
18    Tjonneland A, Gronbaek M, Stripp C, Overvad K. Wine intake and diet in a random sample of
      48763 Danish men and women. Am J Clin Nutr 1999; 69(1): 49-54.
19    Gronbaek M. Factors influencing the relation between alcohol and mortality--with focus on wine. J
      Intern Med 2001; 250(4): 291-308.
20    Ferrari P, Licaj I, Muller DC, Kragh AP, Johansson M, Boeing H e.a. Lifetime alcohol use and
      overall and cause-specific mortality in the European Prospective Investigation into Cancer and
      nutrition (EPIC) study. BMJ Open 2014; 4(7): e005245.
21    Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA, Jr., Stampfer MJ, Willett WC e.a. Roles
      of drinking pattern and type of alcohol consumed in coronary heart disease in men. N Engl J Med
      2003; 348(2): 109-118.
22    Rimm EB, Klatsky A, Grobbee D, Stampfer MJ. Review of moderate alcohol consumption and
      reduced risk of coronary heart disease: is the effect due to beer, wine, or spirits. BMJ 1996;
      312(7033): 731-736.
Pagina 20
</pre>

====================================================================== Einde pagina 156 =================================================================

<br><br>====================================================================== Pagina 157 ======================================================================

<pre>Alcoholhoudende dranken
GEZONDHEIDSRAAD                                                                   Reactie op commentaren
23    Cleophas TJ. Wine, beer and spirits and the risk of myocardial infarction: a systematic review.
      Biomed Pharmacother 1999; 53(9): 417-423.
24    Tolstrup J, Gronbaek M. Alcohol and atherosclerosis: recent insights. Curr Atheroscler Rep 2007;
      9(2): 116-124.
25    Klatsky AL, Armstrong MA, Friedman GD. Red wine, white wine, liquor, beer, and risk for coronary
      artery disease hospitalization. Am J Cardiol 1997; 80(4): 416-420.
26    DiCastelnuovo A., Costanzo S, Bagnardi V, Donati MB, Iacoviello L, de GG. Alcohol dosing and
      total mortality in men and women: an updated meta-analysis of 34 prospective studies. Arch Intern
      Med 2006; 166(22): 2437-2445.
27    Fillmore KM, Stockwell T, Chikritzhs T, Bostrom A, Kerr W. Moderate alcohol use and reduced
      mortality risk: systematic error in prospective studies and new hypotheses. Ann Epidemiol 2007;
      17(5 Suppl): S16-S23.
28    Mukamal KJ, Chiuve SE, Rimm EB. Alcohol consumption and risk for coronary heart disease in
      men with healthy lifestyles. Arch Intern Med 2006; 166(19): 2145-2150.
29    Marmot M, Brunner E. Alcohol and cardiovascular disease: the status of the U shaped curve. BMJ
      1991; 303(6802): 565-568.
30    Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA. Association of alcohol consumption
      with selected cardiovascular disease outcomes: a systematic review and meta-analysis. BMJ 2011;
      342: d671.
31    Klatsky AL, Udaltsova N. Abounding confounding: sick quitters and healthy drinkers. Addiction
      2013; 108(9): 1549-1552.
32    Maclure M. Demonstration of deductive meta-analysis: ethanol intake and risk of myocardial
      infarction. Epidemiol Rev 1993; 15(2): 328-351.
33    Maclure M. Alcohol intake and risk of myocardial infarction. N Engl J Med 1994; 330(17): 1241.
34    Institute of Medicine Committee on Qualifications of Biomarkers and Surrogate Endpoints in
      Chronic Disease, Christine M.Micheel and John R.Ball e. Evaluation of Biomarkers and Surrogate
      Endpoints in Chronic Disease. Washington, USA: National Academy of Sciences; 2010.
35    Beulens JW, Rimm EB, Hu FB, Hendriks HF, Mukamal KJ. Alcohol consumption, mediating
      biomarkers, and risk of type 2 diabetes among middle-aged women. Diabetes Care 2008; 31(10):
      2050-2055.
Pagina 21
</pre>

====================================================================== Einde pagina 157 =================================================================

<br><br>====================================================================== Pagina 158 ======================================================================

<pre>Alcoholhoudende dranken
GEZONDHEIDSRAAD                                                                     Reactie op commentaren
36    Beulens JW, de Zoete EC, Kok FJ, Schaafsma G, Hendriks HF. Effect of moderate alcohol
      consumption on adipokines and insulin sensitivity in lean and overweight men: a diet intervention
      study. Eur J Clin Nutr 2008; 62(9): 1098-1105.
37    Beulens JW, van Loon LJ, Kok FJ, Pelsers M, Bobbert T, Spranger J e.a. The effect of moderate
      alcohol consumption on adiponectin oligomers and muscle oxidative capacity: a human intervention
      study. Diabetologia 2007; 50(7): 1388-1392.
38    Imhof A, Plamper I, Maier S, Trischler G, Koenig W. Effect of drinking on adiponectin in healthy
      men and women: a randomized intervention study of water, ethanol, red wine, and beer with or
      without alcohol. Diabetes Care 2009; 32(6): 1101-1103.
39    Joosten MM, Witkamp RF, Hendriks HF. Alterations in total and high-molecular-weight adiponectin
      after 3 weeks of moderate alcohol consumption in premenopausal women. Metabolism 2011; 60(8):
      1058-1063.
40    Barefoot JC, Gronbaek M, Feaganes JR, McPherson RS, Williams RB, Siegler IC. Alcoholic
      beverage preference, diet, and health habits in the UNC Alumni Heart Study. Am J Clin Nutr 2002;
      76(2): 466-472.
41    Cullmann M, Hilding A, Ostenson CG. Alcohol consumption and risk of pre-diabetes and type 2
      diabetes development in a Swedish population. Diabet Med 2012; 29(4): 441-452.
42    Cao Y, Willett WC, Rimm EB, Stampfer MJ, Giovannucci EL. Light to moderate intake of alcohol,
      drinking patterns, and risk of cancer: results from two prospective US cohort studies. BMJ 2015;
      351: h4238.
43    Hamajima N, Hirose K, Tajima K, Rohan T, Calle EE, Heath CW, Jr. e.a. Alcohol, tobacco and
      breast cancer--collaborative reanalysis of individual data from 53 epidemiological studies, including
      58,515 women with breast cancer and 95,067 women without the disease. Br J Cancer 2002;
      87(11): 1234-1245.
44    Key J, Hodgson S, Omar RZ, Jensen TK, Thompson SG, Boobis AR e.a. Meta-analysis of studies
      of alcohol and breast cancer with consideration of the methodological issues. Cancer Causes
      Control 2006; 17(6): 759-770.
45    World Cancer Research Fund, American Institute for Cancer Research. Continuous update project
      report. Food, nutrition, physical activity, and the prevention of breast cancer. Washington DC:
      American Institute for Cancer Research; 2010.
Pagina 22
</pre>

====================================================================== Einde pagina 158 =================================================================

<br><br>====================================================================== Pagina 159 ======================================================================

<pre>Alcoholhoudende dranken
GEZONDHEIDSRAAD                                                                       Reactie op commentaren
46    Romieu I, Scoccianti C, Chajes V, de BJ, Biessy C, Dossus L e.a. Alcohol intake and breast cancer
      in the European prospective investigation into cancer and nutrition. Int J Cancer 2015; 137(8):
      1921-1930.
47    Tjonneland A, Christensen J, Olsen A, Stripp C, Thomsen BL, Overvad K e.a. Alcohol intake and
      breast cancer risk: the European Prospective Investigation into Cancer and Nutrition (EPIC).
      Cancer Causes Control 2007; 18(4): 361-373.
48    International Agency for Research on Cancer. IARC monographs on the evaluation of carcinogenic
      risk to humans. Volume 96. Alcoholic beverage consumption and ethyl carbamate (urethane). Lyon:
      International Agency for Research on Cancer; 2010.
49    World Cancer Research Fund, American Institute for Cancer Research. Food, nutrition, physical
      activity, and the prevention of cancer: a global perspective. Washington DC: American Institute for
      Cancer Research; 2007.
50    World Cancer Research Fund, American Institute for Cancer Research. Continuous update project
      report. Food, nutrition, physical activity, and the prevention of colorectal cancer. Washington DC:
      American Institute for Cancer Research; 2011.
51    World Cancer Research Fund, American Institute for Cancer Research. Colorectal Cancer
      Systematic Literature Review. Washington DC: American Institute for Cancer Research; 2010.
52    World Cancer Research Fund, American Institute for Cancer Research. Policy and Action for
      Cancer prevention. Food, nutrition, physical activity, and the prevention of cancer: a global
      perspective. Washington DC: American Institute for Cancer Research; 2009.
53    International Agency for Research on Cancer. IARC monographs on the evaluation of carcinogenic
      risk to humans. Volume 44. Alcohol drinking. Lyon: International Agency for Research on Cancer;
      1988.
54    Baan R, Straif K, Grosse Y, Secretan B, El GF, Bouvard V e.a. Carcinogenicity of alcoholic
      beverages. Lancet Oncol 2007; 8(4): 292-293.
55    World Cancer Research Fund, American Institute for Cancer Research. Breast Cancer Systematic
      Literature Review. Washington DC: American Institute for Cancer Research; 2008.
56    World Cancer Research Fund, American Institute for Cancer Research. Continuous update project
      report. Food, nutrition, physical activity, and the prevention of liver cancer. Washington DC:
      American Institute for Cancer Research; 2015.
Pagina 23
</pre>

====================================================================== Einde pagina 159 =================================================================

<br><br>