HindawiPublishingCorporation JournalofOncology Volume2013,ArticleID906495,19pages http://dx.doi.org/10.1155/2013/906495 Review Article Role of Obesity in the Risk of Breast Cancer: Lessons from Anthropometry AminaAmadou,1PierreHainaut,2andIsabelleRomieu1 1NutritionalEpidemiologyGroup,NutritionandMetabolismSection,InternationalAgencyforResearchonCancer, 150CoursAlbertomas,69372LyonCedex08,France 2InternationalPreventionResearchInstitute,95CoursLafayette,69006Lyon,France CorrespondenceshouldbeaddressedtoIsabelleRomieu;[email protected] Received9October2012;Accepted29November2012 AcademicEditor:DagrunEngeset Copyright©2013AminaAmadouetal. is is an open access article distributed under the Creative Commons Attribution License,whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperly cited. Anestimated1.38millionnewcasesofbreastcancer(BC)arediagnosedeachyearinwomenworldwide.Ofthese,themajority are categorized as invasive ductal cell carcinoma. Subgroups of BC are frequently distinguished into �ve “intrinsic” subtypes, namely, luminal A, luminal B, normal-like, HER2-positive, and basal-like subtypes. Epidemiological evidence has shown that anthropometricfactorsareimplicatedinBCdevelopment.Overallconsistentpositiveassociationshavebeenobservedbetween highbodymassindex(BMI)andwaist-to-hipratio(WHR)andtheriskofBCamongpostmenopausalwomen,whilecon�icting resultspersistforpremenopausalBC, bothforBMIandforotheranthropometricparametersaswellasacrossethnicgroups. Furthermore,someevidencesuggeststhatbodysize,bodyshape,andweightgainduringchildhoodoradolescencemayplayarole intheriskofBC.Inthispaper,wedescribetheevidencelinkinganthropometricindicesatdifferentagesandBCrisk,inorder toimproveourunderstandingoftheroleofbodyfatdistributionintheriskofBC,investigatedifferencesintheseassociations accordingtomenopausalstatusandethnicgroups,anddiscussthepotentialbiologicalmechanismslinkingbodysizeandBCrisk. 1.Introduction eincidenceratesofBCshowaheterogeneousdistribu- tion,whileWesterncountriespresentthehighestincidence Breastcancer(BC)isthemostfrequentlydiagnosedcancerin rates, the lowest incidences are observed in low resource womenworldwide,accountingfor23 (1.38million)ofthe countries. BC ranks as the ��h cause of death from cancer totalnewcancercasesin2008[1,2].emajorityofinvasive overall(458,000deaths),butisstillthemostfrequentcauseof breast neoplasms are categorized as%“invasive ductal cell cancerdeathinwomeninbothdeveloping(269,000deaths) carcinoma,nototherwisespeci�ed”(ICD-O8500/3)[3].is and developed regions (189,500 deaths) [1]. Incidence and entityhaslongbeenrecognizedtoincludetumorswithhet- mortality rates have increased in low-incidence countries, erogeneous molecular characteristics, characterized by dis- particularlyinLatinAmerica(LA)andAfricaandespecially tinctpatternsofgeneexpression[4]andofgenomic/genetic amongyoungerwomen.DatafromLAshowthatthehighest alterations[5].SubgroupsofBCarefrequentlydistinguished BC incident rates are among women aged 30–49. In this intoluminal A (estrogen/progesterone-positive),luminal B, area,BCisthemostcommoncanceramongwomen,andthe HER2+,andso-called“triplenegative”subtypes[6].Among leadingcauseofcancermortalitycompoundedbyproblems these subtypes, HER2+ and basal-like subtypes tend to be of access to screening, diagnosis, and treatment [8, 9]. In more common among premenopausal women, as well as Mexicoforexample,BCdeathhasincreased3.6foldbetween in women of African ancestry. Luminal subtypes are more 1995 and 2005. In Africa, a recent study among Gambian commoninpostmenopausalwomenandamongCaucasians women showed an increase of 6.5 in the incidence rate [7]. of BC from 1990 to 2006 [10] compared to 1.5 annually % % 2 JournalofOncology increaseworldwide[11].Asiancountriesarealsorecording indicesatdifferentagesandBCriskinordertoimproveour signi�cant annual increases, for example 2 in Japan and understandingoftheroleofbodyfatdistributionintheriskof 3–5 inChina[11]. BC,toinvestigatedifferencesintheseassociationsaccording e risk factors related to reproductive%factors such as tomenopausalstatusandethnicgroups,andtohighlightthe dela%yedchildbearing,lowerparity,andreducedbreastfeeding potential biological mechanisms linking body size and BC arebecomingmoreprevalentincountriesineconomictran- risk. sitionbutdonotfullyexplaintheincreaseofBCincidence rates.Otherriskfactorssuchaschangesinlifestyle,physical 2.Methods activity,andanthropometryaswellasethnicityandgenetic susceptibility play a role in the development of BC [12– To review the epidemiologic literature on the association 16]. In particular, overweight and obesity have been clearly of overweight, obesity, fat distribution, and BC risk, we associatedwithanincreasedoverallriskofBC. conducted a MEDLINE and PUBMED search including all Overweightandobesityhavebecomemajorpublichealth publications using height, weight, BMI, WC, HC, WHR, challenges throughout the world in both high and low anthropometricfactors,bodyshape,earlylifebodysize,BC, income countries, with over 1 billion overweight and 315 premenopausal,premenopausalBC,case-control,andcohort million obese adults currently estimated worldwide [17, studiesaskeywords.Ourpaperfollowsthepreferredreport- 18]. In the United States (US) for example, obesity is now ingitemsforsystematicreviewsandmeta-analysis(PRISMA) consideredasthesecondleadingcauseofpreventabledeath statementguidelinesdescribedbyLiberatietal.[52].Wethen aer tobacco consumption. Over two-thirds of the adult examinedthereferencesfromtheidenti�edarticles,previous population is overweight, with approximately one-third of review and meta-analysis between anthropometric factors adults and almost 17 of children and adolescents obese andpremenopausalandpostmenopausalBC.eliterature [17,18].Moststudiesontheassociationbetweenobesityand searchincludedallpublicationsfrom1997to2011andrecent BCriskhavebeencond%uctedonCaucasianpopulationsfrom publicationsuptoFebruary2012. Europe,Canada,andtheUS.However,alimitednumberof studiesinotherethnicgroupssuggestsomedifferencesinthe �.�e���t�o��Me�s��e�e�t���d���eo���es�t� prevalenceofobesityandfatdistributionamongwomen.For example studies in the US show that American women of Overweightandobesityarede�nedasabnormalorexcessive AfricanorHispanicoriginaremorelikelytobeobesethan fat accumulation that may impair health [61]. ere are Caucasian women [19]. Given the disparities in prevalence differentwaystomeasureobesity;BMIisthemostcommonly rates of obesity, it is expected that among women, 69 of usedandiscalculatedasweight(kg)dividedbyheight(m ). Caucasians,87 ofAfricanAmericans,and80 ofHispanics eWorldHealthOrganization(WHO)classi�esthedegree willbeoverweightin2015[19]. % 2 of adiposity in terms of the BMI as follows: underweight, Anthropom%etricindices,height,weight,bo%dymassindex less than 18.5; normal, 18.5–24.9; overweight, 25.0–29.9; (BMI),waistcircumference(WC),hipcircumference(HC), obese, more than 30.0kg/m (Table 1(a)). While BMI is or waist-to-hip ratio (WHR), are commonly used as tools usuallyassociatedwithgeneralobesity,WCandWHRhave for assessing overweight/obesity and recently methods for 2 beenusedasmeasuresofcentralorintra-abdominalobesity, evaluating body shape and body size at different ages have de�ned as waist–hip ratio above 0.90 for males and above also been used. Several studies and meta-analyses have 0.85 for females (Table 1(b)). However, there is evidence of examined the associations between anthropometric indices important ethnic variations in body fat distribution, espe- and BC among both pre- and postmenopausal women. ciallyinAsiansascomparedtoCaucasians.InAsianadults, Overall consistent positive associations have been observed differences in body build and body composition result in a between BMI and WHR and the risk of BC among post- differentrelationshipbetweenBMIandbodyfatdistribution menopausalwomen[20–29],whilecon�ictingresultspersist relativetoCaucasians[62–64].InastudyofyoungJapanese forpremenopausalBC,bothforBMIandforotheranthro- and Australian men, Kagawa and colleagues reported that pometric parameters [29–36]. Possible differences in these Japanesemenwereestimatedtohaveanequivalentamount associationscouldberelatedtodifferentialrolesoffatandfat of body fat as the Australian men at BMI values about 1.5 distributiononmetabolismandtotheircontributiontoBC development among pre- and postmenopausal women [31, units lower than those of the Australians (23.5kg/m and 32]. Waist circumference (WC) and WHR, which correlate 28.2kg/m , resp.) [65]. Similarly, studies among Ch2inese with central (abdominal) obesity have been associated with reported s2imilar relationship between BMI and body fat increased BC risk in postmenopausal women [32, 37–39], distribution [43]. ese observations suggest that Asian but not in premenopausal women for whom both null peoplearemorelikelytohavehigherlevelsofbodyfatoverall [23, 38, 39] and increased risk have been reported [31, 32, andmoreabdominalfatandlowerleanmassthantheother 37, 40]. Furthermore, the relation between premenopausal ethnic groups for a given BMI [41–43]. Consequently, the BC and anthropometry has shown inconsistencies across level of BMI that the WHO recommends for overweight ethnic groups [41–43]. ese inconsistencies may be due and obesity in Caucasians may not be applicable for Asian to important ethnic variations in body fat distribution. adults.eWHOWesternPaci�cRegionalO�ce(WPRO) We conducted a systematic review of published studies to has proposed a de�nition of obesity in Asian populations evaluatethestrengthofassociationsbetweenanthropometric withoverweightde�nedasBMI23.0–24.9kg/m andobesity 2 JournalofOncology 3 T1:(a)Combinedrecommendationsofbodymassindexandwaistcircumferencecut-offpointsmadeforoverweightorobesity,and associationwithdiseaserisk.(b)WorldHealthOrganisationcut-offpointsandriskofmetaboliccomplications. (a) Bodymassindex Obesityclass Diseaserisk(relativetonormalweightandwaistcircumference) Men 102cm Men 102cm Women 88cm Women 88cm < > Underweight 18.5 < > Normal 18.5–24.9 < Overweight 25.0–29.9 Increased High 30.0–34.9 I High Veryhigh Obesity 35.0–39.9 II Veryhigh Veryhigh Extremeobesity 40 III Extremelyhigh Extremelyhigh (b) > Indicator Cut-offpoints Riskofmetaboliccomplications Waistcircumference Men 94cm;women 80cm Increased Waistcircumference Men 102cm;women 88cm Substantiallyincreased Waist-hipratio Men >0.90cm;women>0.85cm Substantiallyincreased > > ≥ ≥ Author, year Ethnicity/race RR 95% CI Cohort studies Lahman et al., 2004 [23] Caucasian 1.31 1.08–1.59 Tehard and Clavel-Chapelon, 2006 [28] Caucasian 1.44 1.04–1.99 Palmer et al., 2006 [35] African-American 0.78 0.58–1.05 Kuriyama et al., 2005 [46] Asian 2.67 1.03–6.92 Iwasaki et al., 2007 [47] Asian 2.28 0.94–5.52 Kawai et al., 2010 [22] Asian 2.54 1.16–5.55 Case-control studies Friedenreich et al., 2002 [48] Caucasian 0.99 0.74–1.32 Hall et al., 2000 [40] Caucasian 1.08 0.58–2.01 Berstad et al., 2010 [30] Caucasian 0.75 0.53–1.06 Wenten et al., 2002 [50] Caucasian 2.77 0.86–8.91 Slattery et al., 2007 [36] Caucasian 1.61 1.05–2.46 Hall et al., 2000 [40] African-American 0.68 0.33–1.41 Berstad et al., 2010 [30] African-American 1.26 0.69–2.31 Ogundiran et al., 2010 [34] African 0.76 0.48–1.21 Adebamowo et al., 2003 [49] African 1.82 0.77–4.28 Chow et al., 2005 [20] Asian 1.73 1.04–2.87 Wu et al., 2007 [39] Asian 1.35 0.95–1.93 Mathew et al., 2008 [24] Asian 1.29 1.00–1.67 Wenten et al., 2002 [50] Hispanic 1.32 0.47–3.72 Slattery et al., 2007 [36] Hispanic 0.80 0.44–1.45 Meta-analysis studies Van Den Brandt et al., 2000 [29] Caucasian 1.26 1.08–1.47 Suzuki et al., 2009 [27] All 1.33 1.19–1.48 Renehan et al., 2008 [25] All 1.12 1.05–1.19 0 2 4 6 8 RR F1:ForestplotoftheassociationbetweenBMIandbreastcancerriskinpostmenopausalwomen.esizeofeachboxindicatesthe relativeweightofeachstudy;thehorizontalbarsshowthe95 con�denceintervals(CI).RR:relativerisk;95 CI:95 con�denceintervals. % % % 4 JournalofOncology Variablesofadjustmentorcomment Studycenter,age,educationalattainment,smokingstatus,alcoholconsumption,parity,ageat�rstpregnancy,ageatmenarche,andcurrentpilluse FHBC,ageatmenarche,ageat�rstbirth,parity,historyofbenignbreastdisease,alcoholconsumption,numberofyearsofeducation,maritalstatus,andphysicalactivity Age,educationstatus,ageatmenopause,ageatmenarche,parity,ageat�rstbirth,BMIatage18years,OC,HRT,smoking,alcohol,andphysicalactivitylevel Age,timeperiod,parity/ageat�rstbirth,FHBC,personalhistoryofbenignBC,height,alcoholintake,OCuse,birthweight,ageatmenopause,andHRTuse Age,smoking,alcohol,consumptionofmeat,�sh,fruits,greenoryellowconsumptionofbean-pastesoup,typeofhealth,parity,ageatmenarche,andageatFFTpregnancy Age,area,numberofbirths,ageat�rstbirth,andheight Age,education,smoking,alcohol,andtimespentwalking,Menstrualandreproductivefactors,HRT,andFHBC Age,ageatmenarche,parity,ageat�rstbirth,andfamilyhistoryofbreastcancer Age,totalcaloricintake,physicalactivity,educationallevel,HRT,diagnosedwithbenignBC,FHBC,alcohol,andcurrentsmoke dedinthepaper. RR(95CI) %1.31(1.08–1.59)0.66(0.45–0.98) 0.94(0.74–1.21)0.85(0.60–1.20) 1.44(1.04–1.99)1.03(0.83–1.28) 0.85(0.74–0.98) 0.93(0.90–0.95)0.91(0.89–0.93) 2.67(1.03–6.92) 2.28(0.94–5.53) 2.54(1.16–5.55) 0.55(0.37–0.82)0.78(0.58–1.05)1.01(0.74–1.40) 0.99(0.74–1.32)1.43(1.07–1.93) u cl n i n e menopausalwome Rangeofexposure 30versus2530versus25><><0.84versus0.730.84versus0.73><><30versus18.50.82versus0.74 ≥<≥ boveversusaverag 0.1unitincrease0.1unitincrease 30versus25 ≥< 30versus19 ><25versus20 25versus20≥<37versus250.87versus0.71≥<≥<31.3versus24.1≥<0.83versus0.75 ≥<≥< st A o p n i Y dies 10Y–20 stu ure 12 5–10 racteristicsof Typeofexpos BMInon-HRTHRTWHRnon-HRTHRT BMIWHR Weightatage BodyfatnessChildhoodAdolescent BMI BMI BMI BMIatage18CurrentBMIWHR BMIWHR a h c 2:Selected Cases/controlsorP-years 1,405/103,344 860/41497 2503/35941 4974/188,860 65/9,666 441/55,537 108/10,106 455/59,000 771/762 T Population Caucasian Caucasian Caucasian Caucasian Asian Asian Asian African-American Caucasian n ol udydesig ohort ohort ohort ohort ohort ohort ohort ohort ase-contr St C C C C C C C C C r, 4] 5] 5] Firstauthor,yeaandcountry Lahmann2004Germany[23] Tehard2006France[28] Bardia2008UnitedStates[4 Baer2010UnitedStates[4 Kuriyama2005Japan[46] Iwasaki2007Japan[47] Kawai2010Japan[22] Palmer2007UnitedStates[3 Friedenreich2002Canada[48] JournalofOncology 5 Variablesofadjustmentorcomment Ageatmenarche,nulliparity,breastfeeding,Abortionormiscarriage,BMI,WHR,oralcontraceptive,HRT,FHBC,smoking,alcohol,education,medicalradiationtothechest Age,ageatmenarche,parity/ageatFFTpregnancy,lactation,andeducation Age,race,education,studysite,FHBC,parity,ageatmenopause,andHRT Age,ethnicity,education,menarche,parity,ageatFFTP,breastfeeding,menopausalstatus,ageatmenopause,FHBC,benignbreastdisease,contraceptive,alcohol Age,ageatmenarche,ageatFFTpregnancy,height Age;FHBC;totalMET-hours;parity;OC;breastfeeding;ageat�rstlivebirth; monthsofHRTuse Age,center,physicalactivitylevel,energyintake,alcoholintake,ageatmenopause,parity,andHeight Age,menopausalstatus,ageatmenarche,parity,numberofbirth,ageatFFTbirth,HRT,OC,breastfeeding,smoking,height,weight,andBMI Age,numberofpregnancies,familyhistoryofbreastcancer,income,smoking,alcohol,useofOC,andeducation RR(95CI) % 1.40(1.10–1.70) 0.68(0.33–1.42)1.08(0.58–2.00) 1.62(0.70–3.79)1.64(0.88–3.07) 0.72(0.55–0.96) 0.75(0.53–1.06)1.26(0.55–1.85) 0.76(0.48–1.21) 1.82(0.78–4.31) 2.77(0.86–8.89)1.32(0.47–3.72) 1.61(1.05–2.45)0.80(0.44–1.45) 1.51(0.93–2.46)0.77(0.39–1.50) 8.19(3.40–19.50) 1.73(1.04–2.86)3.82(1.03–14.27) 4.64.6 7777 Continued. Rangeofexposure 0.8versus0.8 >≤ 30.1–59.2versus14.6–230.1–59.2versus14.6–2 0.86–1.34versus0.6–0.0.86–1.34versus0.6–0. 25versus20 ≥<35versus2535versus25≥<≥< 28versus21 ≥< 30versus20 ≥<30versus2230versus22≥<≥<30versus2530versus25><><0.9versus0.80.9versus0.8><><0.86versus0.75 >< 23–27versus1931versus19 <<< 2: n r T ure rs merica Tuser RTuse peofexpos HR MIBlackWhiteHRBlackWhite MIat18yeacentBMICaucasianAfrican-A MI MI MICaucasianHispanic MInon-HR CaucasianHispanicHRnon-HCaucasianHispanic HR MI y W W e W W T B BR B B B B B s ol ntrs 06 25 Cases/coorP-year 436/354 382/419 1,900/2,0 498/266 234/273 687/820 2,325/2,5 130/585 198/353 Population CaucasianAfrican-American CaucasianAfrican-American CaucasianAfrican-American African African CaucasianHispanic CaucasianHispanic Asian Asian n ol ol ol ol ol ol ol ol ol desig contr contr contr contr contr contr contr contr contr y - - - - - - - - - d e e e e e e e e e u as as as as as as as as as St C C C C C C C C C Firstauthor,year,andcountry Huang1999UnitedStates[38] Hall2000UnitedStates[40] Berstad2010UnitedStates[30] Ogundiran2010Nigeria[34] Adebamowo2003Nigeria[49] Wenten2002UnitedStates[50] Slattery2007UnitedStates[36] Ng1997Singapore[33] Chow2005China[20] 6 JournalofOncology Variablesofadjustmentorcomment Age,Asianethnicity,durationofresidenceintheUS,education,menarche,parity,menopausalstatus,ageatmenopause,intakeofteaandsoy,andphysicalactivity Age,center,religion,maritalstatus,educationsocioeconomicstatus,residencestatus,parity,ageat1stchildbirth,durationofbreastfeeding,andphysicalactivity Ameta-analysiswasdonetosummarizetheliteratureonWHRandbreastcancerriskpublishedfromJanuary1966toAugust2002 Lifestylesandreproductivefactors(confoundersthatwerefoundtobesigni�cantinproportionalhazardregressionanalysis)Meta-analysisofcohortandcase-controlstudies(from1970to2007)wasperformedtoclarifytheassociationbetweenbodyweightandtheincidenceofBCde�nedbyER/PRstatusofthetumors HRT,OC,historyofbenignbreastdisease,FHBC,smokingstatus,education,fatintake,�berintake,energyintake,andalcoholintake Cohortandcase-controlstudiespublishedfrom1966toNovember2007wereincludedintheanalysis.edoseresponsemeta-analysiswasadjustedbygeographicregionandcancersite Age,ethnicity,comorbidity,andmenopausalstatus placementtherapy;OC:oralcontraceptives;FFT: e r RR(95CI) %1.35(0.95–1.93)1.48(1.02–2.15) 1.29(1.00–1.66)1.00(0.64–1.54) 1.50(1.10–2.04) 1.32(1.16–1.49) 1.82(0.85–3.85) 1.33(120–1.48) 1.26(1.09–1.46) 1.12(1.08–1.16)1.15(1.08–1.23)1.09(1.04–1.48)1.31(1.15–1.48) 4.8(1.8–12.7)1.4(0.5–4.1) cer;HRT:hormonal n a c ast Continued. Rangeofexposure 24.60versus20.430.84versus0.76 >≤>≤ 25–29.9versus2530versus25 <>< 0.1unitincrease 0.75versus0.80 0.75versus0.80 0.5unitincrease 21versus31 5unitsincrease5unitsincrease5unitsincrease5unitsincrease 30versus2530versus25≥<gesteronereceptor;BC:bre≥< o 2: pr T Typeofexposure BMIWHR BMI WHR WHRCohortstudiesCase-controlstudies BMIER+PR+ BMI BMIallNorthAmericanEuropeanandAustralianAsia-Paci�c BMIAfrican-AmericanHispanic R:estrogenreceptor;PR:ars. Ee Cases/controlsorP-years 1,277/1,160 968/691 19studies 8studies:5cohortand3case-control 31studies:9cohortand22case-control 4,385/337,819 31studies 237/234 R:waist-hipratio;breastcancer;�:y Hof Wy Population Asian Asian All All All Caucasian CaucasianAsian African-AmericanHispanic 2(inkg/m);familyhistor Firstauthor,year,Studydesignandcountry Wu2007Case-controlJapan[39] Mathew2008Case-controlIndia[24] Connolly2002Meta-analysisCanada[37] Harvie2003Meta-analysisUnitedkingdom[32] Suzuki2009Meta-analysisSweden[27] VanDenBrandt2000Meta-analysisUnitedStates[29] Renehan2008UnitedStatesMeta-analysis[25] Sarkissyan2011Cross-UnitedStates[51]sectional BMI:measurementofbodymassindexageat�rstfull-termpregnancy;FHBC:Bold:statisticallysigni�cant. JournalofOncology 7 de�nedasBMI 25kg/m .Inaddition,WHOhassuggested due to the small number of cases, in particular in post- that a waist circumference of 90cm for men and 80cm for menopausal women, re�ecting the demographic structure 2 women was use≥d as interim lower values for Asians as the of African populations. Another possible explanation for cut-off point to de�ne overweight instead of 94cm in men these differences may be the selection bias mainly in case- and85cmforCaucasianwomen[66]. control studies, where anthropometric measurements are oen obtained aer diagnosis and may be affected by the effectsofcancerdevelopmentonthegeneralconditionofthe 4.InPostmenopausalWomen patient. e lack of association between BMI and BC risk observedinHispanicwomenmayre�ectdifferenteffectsof 4.1.BMIandBreastCancerRisk. Mostavailablestudiesand fataccumulationanddistributioninthisHispanicpopulation meta-analyses have focused on BMI as a marker of general comparedtopopulationsinhighresourcecountries. obesity [22, 24, 25, 27–29, 67] and indicated an overall increase in the risk of postmenopausal BC in overweight or obese women among all ethnic groups (Table 2). Figure 4.2.WHRandBreastCancerRisk. WHRiscommonlyused 1 summarizes studies of the association between BMI and as a measure of central obesity [31, 32, 70], de�ned as breast cancer risk in postmenopausal women. A meta- waist-hip ratio above 0.90 for males and above 0.85 for analysis of some of these studies conducted by Renehan femaleshasoenbeenassociatedwiththeriskofdeveloping observed an overall 12 increase per 5kg/m increase in postmenopausalBC(Table2).However,whilemoststudies BMI (RR ; CI: 1.08–1.16) [25]. However estimates have reported a signi�cant increased risk [33, 37–39, 48], 2 may vary according to%geographic regions; the association somestudiesareinconclusive[23,35,40](Figure2).Ameta- tended to b=e s1t.r1o2nger in studies from Asia Paci�c (RR analysiswithsixcase-controland�vecohortstudiesobserved ;95 CI:1.15–1.48)thanstudiesfromNorthAmerica, a summary risk estimate of 1.50 (95 CI: 1.10–2.04) for Europe, and Australia (RR = 1.15; 95 CI: 1.08–1.23 an=d postmenopausal women [37]. e associations tend to be 11..3019;95%CI:1.00–1.14,resp.).Pooleddatafromsevencohort strongerinAsianwomenthanotheret%hnicgroups[33].Ina studies including 337,819 women and%4,385 incident BC Canadiancase-controlstudy,apositiveincreaseinRRof1.43 casesfo%unda26 increaseinpostmenopausalBCriskwith (95 CI:1.07–1.93)wasobservedwhencomparingthehigh- BMIsgreaterorequalto28kg/m [29].Inmostcase-control est ( 0.83) to the lowest quintile ( 0.75) [48]. Similarly, in and prospective%studies, a positi2ve and strong relationship the%Carolinacase-controlstudy,theORforpostmenopausal hasbeenobservedbetweenBMIandpostmenopausalAsian wom>en in the highest quintile ( 0<.8) compared to lowest women [20, 22, 67]. In a cohort study of 10,106 women, quintile ( 0.8) was 1.40 (95 CI: 1.1–1.7) [38]. A study in conductedinJapan,theRRfordevelopingpostmenopausal Africa(Nigeria)reportedanRRof>2.67;95CI(1.05–6.80)for BC was 2.54; 94 CI (1.16–5.55) in women with BMI of postmeno≤pausalwomenwith%WHR 0.85comparedtothose 25kg/m or above compared to those with less 20.5kg/m withWHR 0.77[71].Incontrast,somestudiesconducted [22]. In2contrast%, there is a lack of data among women2 intheUSdidnotdetectasigni�can>tassociation.Halletal. of African origin, but the notion that increase in BMI reportedan<onincreasedRRof1.62;95 CI(0.70–3.79)in is associated with the risk of postmenopausal BC in this AfricanAmericanwomenandof1.64;95 CI(0.88,3.07)for population is consistent with a recent case-control study CaucasianAmericanwomenwhencomp%aringhighestversus reporting a positive association between high BMI and an lowest quintiles (0.86–1.34 versus 0.6–0%.77) [40]. However increasedriskofestrogenreceptorandprogesteronereceptor the power of the study was limited by the small number positiveBC(OR=1.83;95 CI:1.08–3.09;P trend=0.03) of cases (179 cases and 182 controls in African women). RegardingHispanicwomen,onlyonestudyhasassessedthe comparingthehighest( 35kg/m )tothelowest( 25kg/m ) associationbetweenWHRandBCrisk.isstudyfoundno quintile[30].Otherstudies%onwo2menofAfricandescentd2o signi�cant association between WHR and postmenopausal not support a positive>association between BMI<and post- BCrisk[36](Figure2). menopausalBC[34,35,40,49].Inaprospectivestudyamong AfricanAmericans,Palmeretal.reportedanonstatistically signi�cantdecreasingtrendforwomenwithBMIof35kg/m 5.InPremenopausalWomen or above compared to 25kg/m [35]. Similarly, a recen2t case-control study including 505 postmenopausal BC cases 5.1.BMIandBreastCancerRisk. Previousstudies,reviews, 2 and 278 controls found<a RR of 0.76; CI (0.48–1.21) for and pooled analyses have addressed the association of BMI womenwithBMI 28kg/m comparedto 21kg/m [34]. and premenopausal BC (Table 3). Figure 3 summarizes FewstudiesontheassociationbetweenobesityandBChave studies of the association between BMI and breast cancer 2 2 been conducted in≥Hispanics, of those, mos<t have reported risk in premenopausal women. In 2000, Van den Brandt et inconsistent association between BMI and risk of BC in al.analyzedtheassociationbetweenpremenopausalBCand postmenopausalwomen[50,68,69].Recently,Sarkissyanet BMIusingpooleddatafromsevenprospectivecohortstudies al.(2011)reportedinacross-sectionalstudy,alackofassoci- in Caucasian women. Multivariate analyses controlling for ationbetweenobesitymeasuredbyBMIandpostmenopausal reproductive, dietary, and other risk factors detected that BC among Hispanic women in the US population (OR = when compared to premenopausal women with a BMI of 1.4, 95 CI: 0.5, 4.1) [51]. e heterogeneity among these lessthan21kg/m ,womenwithaBMIexceeding31kg/m �ndingsmaybeduetothelackofpowerinAfricanstudies hadanRRof0.54(95 CI:0.34–0.85),suggestinganinverse 2 2 % % 8 JournalofOncology Author, year Ethnicity/race RR 95% CI Cohort studies Lahman et al., 2004 [23] Caucasian 0.94 0.73–1.20 Tehard and Clavel-Chapelon, 2006 [28] Caucasian 1.03 0.83–1.28 Palmer et al., 2006 [35] African-American 1.01 0.74–1.39 Case-control studies Friedenreich et al., 2002 [48] Caucasian 1.43 1.06–1.93 Huang et al., 1999 [38] Caucasian 1.40 1.13–1.73 Slattery et al., 2007 [36] Caucasian 1.51 0.93–2.46 Hall et al., 2000 [40] Caucasian 1.64 0.88–3.07 Hall et al., 2000 [40] African-American 1.62 0.70–3.77 Adebamawo et al., 2003 [71] African-American 2.67 1.05–6.79 Slattery et al., 2007 [36] Caucasian 1.51 0.93–2.46 Ng EH et al., 1997 [33] Asian 8.19 3.42–19.64 Wu et al., 2007 [39] Asian 1.48 1.02–2.15 Slattery et al., 2007 [36] Hispanic 0.77 0.39–1.51 Meta-analysis studies Connolly et al., 2002 [37] All 1.32 1.17–1.49 Harvie et al., 2003 [32] All 1.33 1.19–1.48 0 1 2 3 4 5 6 7 RR F2:ForestplotoftheassociationbetweenWHRandbreastcancerriskinpostmenopausalwomen.esizeofeachboxindicatesthe relativeweightofeachstudy;thehorizontalbarsshowthe95 con�denceintervals(CI).RR:relativerisk;95 CI:95 con�denceintervals. % % % associationbetweenBMIandriskofpremenopausalBC[29]. casesand1,182controlsreportedanincreasedriskof33–56 Adose-responsemeta-analysisonBMIandpremenopausal when comparing premenopausal women above 25kg/m BC was conducted by Renehan et al. using a dataset of 20 with those with than less 25kg/m (OR = 1.33; 95 C%I2: studies. is analysis detected an overall risk estimate of 1.50–1.62forBMI25–29.9kg/m and21.56;95 CI:1.03,2.35 0.92(95 CI:0.88–0.97)foreachincrementof5kg/m [25]. forBMIgreaterthan30kg/m )[224].Interestinglyoner%ecent Several studies supported the hypothesis that higher level 2 cohort study conducted in Japan reported a%high increased of BMI%may be associated with a decrease in the risk of 2 risk of 2.54; 95 CI (1.16–5.55) when comparing women premenopausal BC. is hypothesis is supported by results of BMI 25kg/m to those of 20kg/m [22]. In contrast, fromseveralcase-controlstudies[34,39,40,60]andcohort otherstudiesam%ongAsianwomendidnotdetectasigni�cant studies [35, 53]. However, others studies did not observe a 2 2 statistically signi�cant association when comparing highest associationbetweenBMIandth<eriskofpremenopausalBC [20, 47]. Among African women, the majority of studies versus lowest levels of BMI [23, 49, 56]. Ethnicity appears to modify the association of overweight, obesity, and BC. reported an inverse association between obesity and pre- While the inverse association between BMI and risk of menopausalBC.Acase-controlstudyconductedinNigeria premenopausal BC is well documented in Caucasians, the reportedasigni�cantdecreasingtrendinpremenopausalBC association among Asian women is inconsistent. Several withincreasingBMI(Ptrend=0.027)[34].Asimilar�nding studies among Asian women suggest that higher BMI may wasobservedinalargeprospectivestudyconductedamong be associated with an increased risk for premenopausal BC African Americans in whom the multivariate adjusted RR [22,24,46,55].Aprospectivestudyincluding11,889women was 0.72 (95 CI: 0.54–0.96) for BMI 25kg/m relative from Taiwan reported that higher BMI was moderately to BMI 20 [35]. In contrast, some studies did not �nd a 2 associatedwithanincreasedriskofpremenopausalBC[55], signi�cant as%sociation [30, 40]. With re>spect to Hispanics, with an OR of 1.90 (1.00–3.4) for BMI 26.2kg/m versus a case-co<ntrol study conducted on Hispanic women living 21.6kg/m .Similarlyamulti-centriccase-controlstu2dycon- in the US concluded that high BMI at 30 years of age was ducted am2ong urban and rural women>and including 898 associated with a decrease of the risk of premenopausal JournalofOncology 9 BC (OR = 0.46, 95 CI: 0.25, 0.84 for BMI 30 versus 7.VariationswithMolecularSubtypesof 25kg/m ) [36]. A more recent study has shown that high BreastCancer BMI was2inversely a%ssociated with risk of BC>in Hispanic <population(Ptrend 0.01)[58]. e association between obesity and BC appears to vary according to the molecular subtype of BC, as de�ned by gene expression patterns into luminal A and B, HER2+, < 5.2.WHRandBreastCancerRisk. WHRhasnotbeencon- and triple-negative subtypes. A meta-analysis suggests that sistentlyassociatedwithincreasedBCriskinpremenopausal the association between BMI and BC risk is heterogeneous women, for whom both null [23, 38, 39] and increased accordingtoestrogenreceptor(ER)andprogesteronerecep- risk have been reported [31, 32, 37, 40] (Figure 4, Table tor (PR) status of the tumor. Higher BMI was associated 3). Two meta-analyses [32, 37] have reported that a greater with higher risk for ER+/PR+ tumors in postmenopausal WHR was associated with about 1.5 fold increased risk women(RR=1.82;95 CI:1.55–2.14)buthadaprotective of premenopausal BC. A pooled analysis on seven cohorts effectonER+/PR+tumorsinpremenopausalwomen(RR= and four case-controls reported a summary risk estimate 0.80; 95 CI: 0.70–0.9%2). By contrast, no associations were of 1.79; 95 CI (1.22–2.62) [37] but the strength of the observed for ER PR tumors [27]. In 2011, Harris et al. association varied according to ethnic groups [41–43, 64]. investiga%ted the association between body fat distribution, A large pro%spective cohort study of 11,889 women con- assessedin1993b−ysel−f-reportedWC,HC,andWHR,andthe ducted in Taiwan found that central adiposity re�ected by incidenceofpremenopausalBCintheNurses’HealthStudy hip circumference was a signi�cant predictor of BC [55]. II.Eachofthesebodyfatdistributionmeasuresappearedto Similarly,inaprospectivecase-controlstudyinvolving1,086 bestatisticallysigni�cantlyassociatedwithgreaterincidence Chinese women in Singapore, central obesity as indicated of estrogen receptor (ER) negative BC (RR for ER-negative by a larger WHR was associated with highest risk for BC, BC for the highest versus the lowest quintile of WHR was with the OR being 7.81; (95 CI: 2.8–21.9) comparing the 1.95(95 CI:1.10–3.46;Ptrend=0.01)[31].ese�ndings last ( 0.86) and �rst quintile ( 0.75), whereas BMI did not suggest that body fat distribution may be associated with signi�cantlypredicttherisk%forBC[33].AmongCaucasian anincre%asedriskforER-negativeBCamongpremenopausal wome>n, a case-control study i<ncluding 523 cases and 471 women. controls,adjustedforBMI,reportedanincreaseriskof1.04 Inapooledanalysisoftumormarkerandepidemiological (1.01–1.08)foreachincrementof0.1unitofWHRandarisk riskfactordatafrom35,568invasiveBCcasepatientsfrom of2.44;95 CI(1.17–5.09)whencomparingthe(0.86–1.34) 34studiesparticipatingintheBCAssociationConsortium, versus(0.6–0.77)quintiles[40].Fewstudieshavebeencon- obesity (BMI 30kg/m ) in women 50 years was found ductedon%Africanwomen(AfricanAmericanandAfrican); to be more frequent in ER /PR than in ER+/PR+ tumors amongthose,acase-controlstudyconductedamongAfrican 2 ( ).In≥contrast,obesityinwom≤en 50yearswasless American found a positive association between WHR and premenopausalBC(RR=2.50;95 CI:1.1,5.67,Ptrend frequenti−n7PR thaninPR−+tum−ors( )[75].is 0.005)forwomenwithWHR(0.86–1.34)comparedtothose s𝑃𝑃tud𝑃y𝑃d𝑃𝑃e𝑃monstrated that elevated BMI in>you−n4ger women withWHR(0.6–0.77)[40].Others%tudies[55,58,72]didno<t ( 50years)wa−sassociatedwiththeris𝑃𝑃k𝑃of𝑃E𝑃𝑃R𝑃+orPR+BC �ndastaticallysigni�cantassociation.Overall,thisincreased butnotwithER /PR BCandsuggestedthattriple-negative risk associated with larger WHR among premenopausal o≤rcorebasalphenotype(CBP);de�nedbytriple-negativeand women is found to be stronger amongst Asian women cytokeratins[CK−s]5/−6+(CBP)BCmayhavedistinctetiology comparedtootherethnicgroups.Studiesconductedbetween [75]. e differences between the conclusions of the two WHR and premenopausal BC among Hispanic women are studies[31]maybelinkedtotheuseofdifferentobesity/fat limited and none has reported signi�cant association. For distribution measures. Further studies are needed to better example,Johnetal.,reportedanonsigni�cantdecreaserisk understandtheassociationsbetweenoverweightandobesity ofhavinggreaterWHRamongHispanicwomen(OR=0.71, andspeci�csubtypesofBC. 95 CI:0.46–1.1;Ptrend=0.19)[58]. 8.EffectofEarlyLifeAnthropometryon % �.Si�ni�canceof�aist�to��ei�htRatio BreastCancerRisk ere is evidence suggesting that the waist-to-height ratio Mostoftheepidemiologicalstudieshavefocusedmainlyon (WHtR)maybeamoreusefulglobalclinicalscreeningtool adult BMI and not on weight change or on the in�uence thanWC,withaweightedmeanboundaryvalueof0.5[73]. ofearlylifebodyweightandbodyshape(silhouette).ese e WHtR is de�ned as waist (cm) circumference divided factors, such as birth size, body shape, and weight during by height (cm) and correlates well with abdominal obesity; childhoodoradolescencemayplayaroleintheriskofBC. highervaluesofWHtRindicatehigherriskofobesity-related Epidemiologicstudieshavedemonstratedthatindependently cardiovasculardiseases[74].However,todatetheassociation ofBMI,greaterbodyfatnessduringchildhoodoradolescence between WHtR and BC risk has not been systematically are associated with lower BC risk in both premenopausal studied.Nevertheless,apreviousstudyofmultiethnicgroups women[30,36,45,53,54,57]andpostmenopausalwomen reportedaninverseassociationbetweenWHtRandER+PR+ [30, 35, 44, 45]. In a recent large prospective study con- BCinallpremenopausalwomen[58]. ductedamong188,860women,Baerandcolleaguesreported 10 JournalofOncology Variablesofadjustmentorcomments Ageatenrolment,parity,ageatFFTpregnancy,OC,ageatmenarche,FHBC,totaldurationofbreastfeeding,andcountry Studycenter,age,educationalattainment,smoking,alcohol,parity,ageat�rstpregnancy,ageatmenarche,andcurrentpilluse Age,timeperiod,birthweight,height,recentalcoholconsumption,parity,ageat�rstbirth,OC,historyofbenignbreastdisease,and�rstdegreeofFHBC FHBC,ageatmenarche,ageatFFTP,parity,historyofbenignbreastdisease,alcoholconsumption,education,maritalstatus,andphysicalactivity Age,timeperiod,parity,ageat�rstbirth,FHBC,personalhistoryofbenignbreastdisease,height,alcoholintake,OC,andbirthweight Age,height,historyofbenignbreastdisease,FHBC,ageatmenarche,ageatFFTP,parity,OC,alcohol,andphysicalactivity Age,smoking,alcohol,consumptionofmeat,�sh,fruits,greenoryellowconsumptionofbean-pastesoup,typeofhealth,parity,ageatmenarche,andageatFFTpregnancy Age,area,numberofbirths,ageat�rstbirth,andheight Ageatenrollment,height,weight paper. CI) %40–0.97)59–0.91)50–0.93) 59–1.14)74–1.50) 61–0.98)65–0.91) 67–0.99) 06–1.00)39–0.91) 87–0.94) 87–0.94) 10–3.46) 24–2.88)03–6.92) 53–3.47) 00–3.40)30–1.20) e 5 0.0.0. 0.0. 0.0. 0. 0.0. 0. 0. 1. 0.1. 0. 1.0. h 9 ((( (( (( ( (( ( ( ( (( ( (( t ( 249 25 77 2 60 1 8 5 47 7 00 n R 676 80 77 8 26 9 8 9 86 4 96 di R 0.0.0. 0.1. 0.0. 0. 0.0. 0. 0. 1. 0.2. 1. 1.0. nclude erage angeange ange menopausalwomeni Rangeofexposure 30versus2025versus20–24.9Fat/veryfatversusav><≥ 28.8versus21.50.846versus0.736 ≥<>< 2levelsversusnoch2levelsversusnoch≥2levelsversusnoch≥ ≥30versus18.50.82versus0.74 ><≥ 0.2unitincrease 0.1unitincrease 0.84versus0.73 ≥< 29.9versus2530versus25 ≥<≥< 30versus19 26.2versus21.6><0.85versus0.77 ><< e diesinpr ure age7 fatnessto10to20 0to20 5–10Y 10–20Y cteristicsofstu Typeofexpos BMIBMIatage18Bodyshapeat BMIWHR IncreasebodyFromage5Fromage5 Fromage1 BMIWHR BodyfatnessChildhood Adolescent WHRER − BMI BMI BMIWHR a 3:Selectedchar Cases/controlsorP-years 733/99,717 474/73,542 1,318/109,267 275/20,068 2,188/188,860 620/45,799 33/5,214 441/55,537 104/11,889 T opulation aucasian aucasian aucasian aucasian aucasian aucasian sian sian sian P C C C C C C A A A n g si e yd ort ort ort ort ort ort ort ort ort d h h h h h h h h h u o o o o o o o o o St C C C C C C C C C 4 6 Authors,years,andcountry Weiderpassetal.,200Norway-Sweden[53] Lahmannetal.,2004Germany[23] Baeretal.,2005UnitedStates[54] TehardandClavel-Chapelon,200France[28] Baeretal.,2010UnitedStates[45] Harrisetal.,2011UnitedStates[31] Kuriyamaetal.,2005Japan[46] Iwasakietal.,2007Japan[47] Wuetal.,2006Taiwan[55]
Description: