Mendelian randomisation analysis strongly implicates adiposity with risk of developing colorectal cancer

Mendelian andomisa ion analysis s ongly implica es adiposi y wi h isk o de eloping colo ec al cance Da id Ja is 1,34 , Jona han S Mi chell 1,34 ,PhilipJLaw 1 , Kimmo Palin 2,3 , Sa i Tuupanen 2,3 , Alexand a Gyl e 2,3,35 , Ul ika A Ha ¨nninen 2,3 , Ta iana Cajuso 2,3 , Tomas Tanskanen 2,3 , Johanna Kondelin 2,3 , Ee i Kaasinen 2,3 , An i-Pekka Sa in 4 , Jaakko Kap io 4,5 , Johan G E iksson 5,6,7 , Ha i Rissanen 5 , Paul Knek 5 , Ee o Pukkala 8,9 , Pekka Jousilah i 5 , Veikko Salomaa 5 , Samuli Ripa i 4 , Aa no Palo ie 4,10,11,12 , Heikki Ja ¨ inen 13 , Lau a Renkonen-Sinisalo 14 ,AnnaLepis o ¨ 14 ,JanBo ¨hm 15 , Jukka-Pekka Mecklin 16 , Nada A Al-Tassan 17 , Clai e Palles 18 , Lynn Ma in 18 , Ella Ba clay 18 , Susan M Fa ing on 19,20 , Ma ia N Timo ee a 19 , B ian F Meye 17 , Salma M Wakil 17 , Ha y Campbell 21 , Ch is ophe G Smi h 22 , Shelley Idziaszczyk 22 , Timo hy S Maughan 23 , Richa d Kaplan 24 , Rachel Ke 25 , Da id Ke 26 , Daniel D Buchanan 27,28 , Aung K Win 28 , John L Hoppe 28 , Ma k A Jenkins 28 , No alane M Lindo 29 ,PollyANewcomb 30 , S e e Gallinge 31 , Da id Con i 32 , F ed Schumache 32 , G aham Casey 32 , Jussi Taipale 2,3,33 , Lau i A Aal onen 2,3 , Je emy P Cheadle 22 , Malcolm G Dunlop 19 , Ian P Tomlinson 18 and Richa d S Houls on* ,1 Backg ound: Obse a ional s udies ha e associa ed adiposi y wi h an inc eased isk o colo ec al cance (CRC). Howe e , such s udies do no es ablish a causal ela ionship. To minimise bias om con ounding we pe o med a Mendelian andomisa ion (MR) analysis o examine he ela ionship be ween adiposi y and CRC. Me hods: We used SNPs associa ed wi h adul body mass index (BMI), wais -hip a io (WHR), childhood obesi y and bi h weigh as ins umen al a iables in a MR analysis o 9254 CRC cases and 18 386 con ols. Resul s: In he MR analysis, he odds a ios (ORs) o CRC isk pe uni inc ease in BMI, WHR and childhood obesi y we e 1.23 (95% CI: 1.02–1.49, P¼0.033), 1.59 (95% CI: 1.08–2.34, P¼0.019) and 1.07 (95% CI: 1.03–1.13, P¼0.018), espec i ely. The e was no e idence o associa ion be ween bi h weigh and CRC (OR ¼1.22, 95% CI: 0.89–1.67, P¼0.22). Combining hese da a wi h a concu en MR-based analysis o BMI and WHR wi h CRC isk ( o alling o 18 190 cases, 27 617 con ols) p o ided inc eased suppo , ORs o BMI and WHR we e 1.26 (95% CI: 1.10–1.44, P¼7.7 10 4 ) and 1.40 (95% CI: 1.14–1.72, P¼1.2 10 3 ), espec i ely. Conclusions: These da a p o ide u he e idence o a s ong causal ela ionship be ween adiposi y and he isk o de eloping CRC highligh ing he u gen need o p e en ion and ea men o adiposi y. Colo ec al cance (CRC) is a majo public heal h p oblem in economically de eloped coun ies (Fo man e al, 2014). Mig a ion and epidemiological s udies ha e es ablished ha die and o he li es yle ac o s ha e a majo ole in he ae iology o CRC (Hagga and Boushey, 2009). Among he li es yle ac o s ha ha e an impac on CRC isk, high body mass index (BMI) has gene ally *Co espondence: P o esso RS Houls on; E-mail: [email p o ec ed] 34 These au ho s con ibu ed equally o his wo k. 35 Cu en add ess: Human Longe i y Inc., La Jolla, CA 92121, USA. Recei ed 29 Janua y 2016; e ised 9 May 2016; accep ed 14 May 2016; published online 23 June 2016 &2016 Cance Resea ch UK. All igh s ese ed 0007 – 0920/16 FULL PAPER Keywo ds: Mendelian andomisa ion; adiposi y; colo ec al cance B i ish Jou nal o Cance (2016) 115, 266–272 | doi: 10.1038/bjc.2016.188 266 www.bjcance .com | DOI:10.1038/bjc.2016.188 been epo ed o be associa ed wi h an inc eased isk in obse a ional s udies (Bianchini e al, 2002; Ba dou e al, 2013), wi h some e idence o a s onge in luence being shown o men (Pan e al, 2004; Moghaddam e al, 2007). The e is also e idence o an in e se ela ionship be ween adiposi y a a young age and cance de elopmen (Ca pen e e al, 2003; Yang e al, 2014). The associa ion be ween BMI and CRC in hese obse a ional s udies does no , howe e , necessa ily es ablish a causal ela ion- ship. Speci ically, hese s udies canno en i ely exclude he possibili y o he obse ed associa ion being he consequence o con ounding ac o s, such as socio-economic s a us, alcohol and o he li es yle ac o s (Moghaddam e al, 2007), whe eas some s udies ha e ailed o exclude he possibili y o e e se causa ion (Smi h and Eb ahim, 2003). An al e na i e o a adi ional obse a ional epidemiology s udy is he Mendelian andomisa ion (MR) app oach. The s a egy o MR uses gene ic ma ke s known o be associa ed wi h a po en ial isk ac o in he assessmen o i s e ec on ano he ai o disease (Lawlo e al, 2008). These ma ke s, o ins umen al a iables (IV), ely on a numbe o assump ions, namely ha he IVs a e solely associa ed wi h he ai o disease, and he IVs a e independen o con ounde s. This me hodology pe mi s he na u e o he ela ion be ween he isk ac o and he ai o disease o be assessed wi hou he limi a ions p esen wi hin obse a ional s udies, such as con ounding ac o s, and impo an ly es ablish whe he an associa ion is causal. A u he a ibu e o MR is he a oidance o he in luence o ac o s whose e ec may be ime sensi i e, o example weigh loss being a consequence o CRC. Thus, an IV has he po en ial o mo e accu a ely assess li e ime exposu e when compa ed wi h measu e- men s o po en ial isk ac o s eco ded in an obse a ional s udy. In he p esen s udy we examined, using MR, he impac o ou me ics o adiposi y on he isk o de eloping CRC. The adiposi y ai s we conside ed we e adul BMI, adul wais -hip a io (WHR), childhood obesi y and bi h weigh . MATERIALS AND METHODS CRC GWAS da a se s. Ou MR analysis was based on da a om se en p e iously epo ed genome-wide associa ion s udies (GWAS) o CRC (O lando e al, 2016). B ie ly, hese GWAS we e all based on indi iduals wi h Eu opean ances y and comp ise: CCFR1 (1290 cases, 1055 con ols), CCFR2 (796 cases, 2236 con ols), COIN (2244 cases, 2162 con ols), FINLAND (1172 cases, 8266 con ols), UK1 (940 cases, 965 con ols), Sco land1 (1012 cases, 1012 con ols) and VQ58 (1800 cases, 2690 con ols). De ails o he geno yping, quali y con ol and impu a ion o un yped single-nucleo ide polymo phisms (SNPs) geno ypes ha e been p e iously published. Summa y s a is ics om he GWAS we e used o calcula e he a io es ima es o he adiposi y- ela ed SNPs. Ins umen al a iables. Fo each o ou adiposi y ai s (adul BMI, adul WHR, childhood obesi y and bi h weigh ), we used da a om ecen GWAS o indi iduals o Eu opean-decen o each ai . Speci ically o adul BMI, we used da a om he Gene ic In es iga ion o An h opome ic T ai s (GIANT) con- so ium, which was based on an analysis o up o 339 224 indi iduals (Locke e al, 2015). Fo he WHR, da a we e ob ained om a me a-analysis comp ising 61 s udies, made up o up o 190 803 indi iduals (Heid e al, 2010). Fo childhood obesi y, da a was ob ained om 2480 child en wi h ex eme obesi y and 7370 con ols (Wheele e al, 2013). Bi h weigh da a was ob ained om up o 69 308 indi iduals om 43 s udies (Ho ikoshi e al, 2013). We used SNPs ha we e decla ed genome-wide signi ican (i.e., Pp5.0 10 8 ) in hese GWAS as IVs. F om he 97 adul BMI- associa ed SNPs (kg m 2 ) as iden i ied by Locke e al (2015), we con ined he analysis o 76 SNPs ha we e ound in Eu opean popula ions, and we e sepa a ed by a leas 500 kb. In addi ion, we used 14 SNPs o WHR as epo ed by Heid e al (2010); nine SNPs epo ed by Wheele e al (2013), which we e associa ed wi h childhood obesi y (43 s.d. om he mean o he BMI dis ibu ion); and he se en SNPs epo ed by Ho ikoshi e al (2013), which we e associa ed wi h bi h weigh (kg) (Supplemen a y Table 1). None o hese s udies epo ed non- addi i e e ec s o he SNPs on he adiposi y ai , hence pe allele e ec s we e conside ed addi i e. Fo each o he ou adiposi y ai s we ex ac ed he e ec es ima es and associa ed P- alues o each SNP om he se en CRC GWAS. S a is ical analysis. We pe o med MR analysis o assess he associa ion be ween each adiposi y ai and CRC using summa y s a is ics om he CRC GWAS, and he published e ec size o he adiposi y- ela ed ai on CRC. As pe Bu gess e al (2015) he combined a io es ima e ð^ bÞo all SNPs associa ed wi h a pa icula adiposi y ai on CRC isk was calcula ed unde a ixed-e ec s model: ^ b¼PkXkYks2 Yk PkX2 ks2 Yk : X k co esponds o he associa ion be ween SNP kwi h he adiposi y ai and Y k is he associa ion be ween SNP kand CRC isk wi h s anda d e o sYk. The s anda d e o o he combined a io es ima e is app oxima ely gi en by seð^ bÞ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 PkX2 ks2 Yk s We used MR-Egge and in e se weigh ed a iance eg ession o examine o iola ion (e.g., om pleio opy) o he s anda d IV assump ions in ou analysis (Bowden e al, 2015). All analyses we e pe o med using R so wa e (R De elopmen Co e Team, Vienna, Aus ia) and we conside ed a P- alue o o0.05 as signi ican . Me a-analysis wi h published s udies. An o e lapping concu en s udy has ecen ly been published which has also epo ed an MR analysis o es ima e he causal associa ion be ween adiposi y and CRC isk (Th i e al, 2015). The s udy was based on he analysis o 10 226 CRC cases and 10 286 con ols om 11 GWAS, including he CCFR1 s udy, epo ing on he impac o adul BMI and WHR on CRC. The gene ic isk sco e o hese wo adiposi y ai s in his s udy and in ou analysis was based on he same se o SNPs. Hence, o enhance ou powe o es ablish a ela ionship be ween gene ically de ined adiposi y and CRC isk, we pe o med a me a- analysis o ou s udy and his published s udy pooling summa y es ima es o e ec size unde a ixed-e ec s model, a oiding duplica ion o he CCFR1 s udy. RESULTS In six o he se en CRC GWAS he e was a posi i e ela ionship be ween BMI-inc easing alleles and CRC isk (Figu e 1). F om he pooled analysis we iden i ied an odds a io (OR) o 1.23 in isk o CRC pe kg m 2 inc ease o BMI (95% con idence in e al (CI): 1.02–1.49, P¼0.033, es o he e ogenei y be ween s udies I 2 ¼0%, P he ¼0.70). Al hough he ela ionship be ween WHR and childhood obesi y wi h CRC was less consis en be ween s udies han o BMI in he pooled analysis, we also iden i ied a co ela ion be ween bo h adiposi y ai s and CRC isk (Figu e 1). Fo each uni inc ease in WHR we obse ed an OR o 1.59 in CRC isk (95% CI: 1.08–2.34, P¼0.019 es o he e ogenei y ac oss s udies I 2 ¼45%, P he ¼0.09). Fo childhood obesi y he OR o CRC was 1.07 Mendelian andomisa ion o adiposi y and CRC BRITISH JOURNAL OF CANCER www.bjcance .com | DOI:10.1038/bjc.2016.188 267 (95% CI: 1.01–1.13, P¼0.018, es o he e ogenei y be ween s udies I 2 ¼0%, P he ¼0.61). In con as o he ela ionship be ween BMI, WHR, and childhood obesi y and CRC we obse ed no associa ion wi h bi h weigh and isk (P¼0.22, Figu e 1). Using bo h MR-Egge and IVW eg ession es s we did no de ec iola ion o he s anda d IV assump ions in ou MR analysis o BMI, WHR, childhood obesi y o bi h weigh o CRC isk (Supplemen a y Figu e 1, Supplemen a y Table 2). The s onges epo ed SNP associa ions o BMI a e p o ided by s1558902 (FTO) and s13021737 (TMEM18), which ha e a well-es ablished impac on obesi y (F ayling e al, 2007; Rohde e al, 2014). To examine i ou indings o a co ela ion be ween BMI and CRC isk we e p ima ily d i en by hese a ian s we pe o med a sensi i i y analysis excluding hese SNPs. The MR esul s emain s a is ically signi ican , albei sligh ly less p o ound (OR ¼1.24, 95% CI: 1.01–1.51, P¼0.035). The associa ion be ween each o he adiposi y- ela ed SNP and CRC isk is shown in Supplemen a y Figu e 1. Gi en he e is co ela ion be ween measu es o adiposi y (Se dula e al, 1993) we examined he speci ici y o each ai on CRC isk by epea ing ou analysis omi ing SNPs i he e we e o e lapping loci. Wi h he excep ion o adul BMI and childhood obesi y which sha e ou associa ed loci FTO ( s1558902, s1421085), MC4R ( s6567160, s476828), TMEM18 ( s13021737, s12463617) and NEGR1 ( s3101336) all o he adiposi y- ela ed SNPs a e dis inc . Omi ing hese o e lapping SNPs om ou MR analysis s ill p o ided e idence o a co ela ion be ween adul BMI and childhood obesi y wi h CRC isk, albei less signi ican han be o e (OR ¼1.23, 95% CI: 1.00–1.51, P¼0.049 and OR ¼1.10, 95% CI: 1.02–1.20, P¼0.019, espec i ely). To explo e i he e we e gende -speci ic associa ions o CRC o each o he ou adiposi y ai s, we pe o med a s a i ied analysis o ou da a se . Al hough an inc easing numbe o isk alleles o each ai was associa ed wi h an inc eased CRC in bo h men and women, only he ela ionship be ween adul WHR and CRC in men emained s a is ically signi ican (Supplemen a y Table 3; OR ¼2.13, 95% CI: 1.18–3.87, P¼0.013). In he me a-analysis combining hese esul s wi h he da a om Th i e al (2015), adul BMI was associa ed wi h an OR o 1.26 o CRC isk (95% CI: 1.10–1.44, P¼7.7 10 4 ) and adul WHR an OR o 1.40 o CRC isk (95% CI: 1.14–1.72, P¼1.2 10 3 ; Table 1). Al hough he associa ion be ween gene ically in luenced adul BMI was only signi ican in women (ORs in emales and males we e 1.43, 95% CI: 1.17–1.74, P¼4.3 10 4 and 1.12, 95% CI: 0.92–1.38, P¼0.26, espec i ely) and he WHR associa ion was only signi ican o men (ORs in males and emales we e 1.63, 95% CI: 1.20–2.22, P¼2.0 10 3 and 1.22, 95% CI: 0.91–1.63, P¼0.18, espec i ely) hese di e ences we e no s a is ically di e en (P- alues o di e ence we e 0.09 and 0.18, espec i ely). DISCUSSION In his s udy we ha e shown co ela ions be ween IVs o adul BMI, WHR and childhood obesi y, and CRC isk. E en adjus ing o mul iple es ing he co ela ions be ween IVs o adul BMI and WHR emained signi ican . The absence o da a on childhood obesi y in he s udy epo ed by Th i e al (2015), p ecluded me a-analysis o his me ic o adiposi y. We did no iden i y a ela ionship be ween he IVs o bi h weigh and CRC. Al hough his is less likely o be a de e minan o CRC isk pe se,we acknowledge ha ou powe o demons a e a ela ionship was limi ed. Indeed, e en he IVs o adul BMI we used explain B3% o he o al pheno ypic a ia ion in BMI (Locke e al, 2015). S udy ABMI WHR BW OR (95% CI) 1.62 (0.90–2.92) 1.22 (0.78–1.90) 1.60 (0.84–3.05) 1.36 (0.89–2.07) 1.23 (1.02–1.49) 1.22 (0.66–2.27) 1.17 (0.66–2.06) Phe = 0.7 l2 = 0 0 0.5 1 2 2.5 31.5 OR P COIN FIN UK1 VQ58 Me a Sco land1 CCFR1 CCFR2 C 1.11 (0.94–1.31) OR (95% CI) 1.01 (0.86–1.18) 1.05 (0.93–1.19) 1.18 (1.05–1.33) 1.07 (1.01–1.13) 1.03 (0.86–1.22) 1.03 (0.86–1.24) 1.02 (0.90–1.15) CO OR 0.9 1 1.1 1.2 1.3 0.11 0.59 0.60 0.38 0.16 0.52 0.15 0.033 0.21 0.0064 0.018 0.78 0.71 0.39 0.81 0.93 Phe = 0.61 l2 = 0 PS udy COIN FIN UK1 VQ58 Me a Sco land1 CCFR1 CCFR2 D 1.94 (0.76–4.97) 0.89 (0.43–1.81) 0.98 (0.48–2.03) 1.63 (0.57–4.65) 1.56 (0.56–4.32) 0.92 (0.47–1.82) 1.22 (0.89–1.67) 2.16 (0.83–5.67) OR (95% CI) Phe = 0.6 l2 = 0 PS udy OR 012345 COIN FIN UK1 VQ58 Me a Sco land1 CCFR1 CCFR2 B OR 02468101214 0.12 0.17 0.74 0.96 0.37 0.39 0.81 0.22 0.190 0.860 0.019 0.008 0.019 0.950 0.360 0.260 2.21 (0.67–7.24) 1.11 (0.35–3.48) 0.60 (0.25–1.45) 1.04 (0.28–3.84) 3.16 (1.35–7.40) 1.59 (1.08–2.34) 4.44 (1.28–15.39) 1.51 (0.63–3.62) OR (95% CI) Phe = 0.09 l2 = 45 PS udy COIN FIN UK1 VQ58 Me a Sco land1 CCFR1 CCFR2 0.89 (0.57–1.37) Figu e 1. MR esul s o adiposi y ai s on CRC isk om se en GWAS s udies and he me a-analysis. (A) Adul BMI; (B) adul WHR; (C) childhood obesi y (CO); and (D) bi h weigh (BW). Boxes deno e OR poin es ima es, wi h hei a eas p opo ional o he in e se a iance weigh o he es ima e. Ho izon al lines ep esen 95% CIs. The diamond ep esen s he summa y O s, compu ed unde a ixed-e ec s model, wi h 95% CI gi en by he wid h o he diamond. The unb oken e ical line is a he null alue (OR ¼1.0). BRITISH JOURNAL OF CANCER Mendelian andomisa ion o adiposi y and CRC 268 www.bjcance .com | DOI:10.1038/bjc.2016.188 A possible explana ion o a ailu e o demons a e a signi ican co ela ion be ween gene ic BMI and CRC isk could be ha he dis ibu ion o body a is a mo e impo an p edic o o CRC isk in men a he han o al body adiposi y. Suppo o his pos ula e is ha in men WHR has been epo ed o be supe io in p edic ing CRC han BMI in obse a ional s udies. In ou MR-based analysis we indeed ound ha WHR was associa ed wi h CRC in men bu no in women. Gi en ha he gene ic isk sco e o WHR is de i ed om SNPs associa ed wi h WHR, which a e adjus ed o BMI, sugges s ha a dis ibu ion may be impo an o CRC isk o men, whe eas o e all obesi y is mo e impo an o CRC isk o women. The e is e idence ha adiposi y may ha e a mo e signi ican e ec on he de elopmen o colon a he han ec al cance . Gi en ha he landscape o colonic and ec al cance s show di e ences ela ing IVs o molecula ea u es is likely o be in o ma i e in e ms o unde s anding disease ae iology (Yamauchi e al, 2012). Un o una ely he da a se s on which ou s udy has been based did no enable such analysis o be pe o med. An impo an s eng h o ou analysis is ha by implemen ing an MR-based analysis we ha e a oided he po en ial biases in obse a ional s udies o ac o s such as ecall bias and con ound- ing. The combined a io es ima es o he impac o adiposi y on CRC isk hold p o ided he ma ke is independen o ac o s ha may con ound he ela ionship be ween adiposi y and CRC. The indings om ou analysis a e howe e elian on a numbe o key assump ions. Fi s , ha he IVs a e solely associa ed wi h CRC h ough i s associa ion wi h adiposi y a he han pleio opism, which would be seen by a depa u e om linea i y o he ela ionship be ween SNPs and hei e ec size o adiposi y and CRC. We did no obse e such a ela ionship be ween CRC and adiposi y isk SNPs. Second, he IV is independen o ac o s ha con ound obse a ional associa ions. To da e he e is cu en ly no e idence ha he IVs we used a e associa ed wi h ac o s ha migh con ound adiposi y CRC associa ions in con en ional analyses. The gene a ion o subs an i e IVs o highly complex ai s is a majo limi a ion o he MR-based s a egy o in es iga e he ae iological basis o diseases like cance . Accep ing hese ca ea s ou s udy indings can be iewed as quan i ying he causal e ec o adiposi y on CRC isk. Mo eo e , hey gene ally suppo p e iously published obse a ional s udies and p o ide u he e idence o adiposi y being a majo isk ac o o he de elopmen o CRC. In he pooled analysis, gene ically in luenced BMI showed only a s a is ically signi ican associa ion wi h CRC in women. This inding is disco dan wi h obse a ional s udies which ha e gene ally ound a s onge co ela ion in men. I is possible ha some es ima es om obse a ional s udies may ha e been biased owa d he null i hea ie women unde - epo hei weigh . Indeed i is no ewo hy ha his speci ic inding om he me a- analysis is p ima ily d i en by da a om Th i e al (2015), hence independen eplica ion is equi ed. The biological mechanism by which adiposi y inc eases CRC isk emains o be es ablished and se e al mechanisms ha e been a iously sugges ed as explaining he co ela ion. These include inc eased insulin and insulin-like g ow h ac o signalling, ch onic in lamma ion and signalling ia adipokines, such as lep in. Fu he mo e, i is plausible ha inc eased adiposi y may al e he in es inal mic obiome, con ibu ing o gas oin es inal ca cinogen- esis (Ha iss e al, 2009; Kan and Hull, 2011; Aleman e al, 2014). I espec i e o he exac unc ional basis o he associa ion be ween adiposi y and CRC isk, demons a ing ha i is causal makes obesi y an impo an a ge o p ima y p e en ion o CRC in he popula ion. ACKNOWLEDGEMENTS We a e g a e ul o all indi iduals who pa icipa ed in he a ious s udies. This s udy made use o geno yping da a om he 1958 Bi h Coho , kindly made a ailable by he Wellcome T us Case Con ol Conso ium 2. A ull lis o he in es iga o s who con ibu ed o he gene a ion o he da a is a ailable a h p:// www.w ccc.o g.uk/. A he Ins i u e o Cance Resea ch, his wo k was suppo ed by Cance Resea ch UK (C1298/A8362–Bobby Moo e Fund o Cance Resea ch UK). Addi ional suppo was p o ided by he Na ional Cance Resea ch Ne wo k. DJ was suppo ed by a summe s uden g an om he BBRSC. In Edinbu gh, he wo k was suppo ed by P og amme G an unding om Cance Resea ch UK (C348/A12076). In Ox o d, addi ional unding was p o ided by he Ox o d Comp ehensi e Biomedical Resea ch Cen e and he EU FP7 CHIBCHA g an . Co e in as uc u e suppo o he Wellcome T us Cen e o Human Gene ics, Ox o d was p o ided by g an (090532/Z/09/Z). We a e g a e ul o many colleagues wi hin UK Clinical Gene ics Depa - men s ( o CORGI) and o many collabo a o s who pa icipa ed in he VICTOR and QUASAR2 ials. We also hank colleagues om he UK Na ional Cance Resea ch Ne wo k ( o NSCCG). Suppo om he Eu opean Union (FP7/207-2013, g an 258236) and FP7 collabo a i e p ojec SYSCOL and COST Ac ion in he UK is also acknowledged (BM1206). The COIN and COIN-B ials we e unded by Cance Resea ch UK and he Medical Resea ch Council and we e conduc ed wi h he suppo o he Na ional Ins i u e o Heal h Resea ch Cance Resea ch Ne wo k. COIN and COIN-B ansla ional s udies we e suppo ed by he Bobby Moo e Fund om Cance Resea ch UK, Teno us, he Kidani T us , Cance Resea ch Wales and he Na ional Ins i u e o Social Ca e and Heal h Resea ch Cance Gene ics Biomedical Resea ch Uni (2011–2014). NAA, BFM and SMW we e unded and suppo ed by KFSHRC. In Finland, his wo k was suppo ed by g an s om he Academy o Finland (Finnish Cen e o Excellence P og am 2012–2017, 250345), he Jane and Aa os E kko Founda ion, he Finnish Cance Socie y ( o KP), he Eu opean Resea ch Council (ERC; 268648), he Sig id Juselius Founda ion, SYSCOL, he No dic In o ma ion o Ac ion eScience Cen e (NIASC), he No dic Cen e o Excellence inanced by No dFo sk (p ojec 62721, o KP) and S a e Resea ch Funding o Kuopio Uni e si y Hospi al (B1401). We acknowledge he compu a ional esou ces p o ided by he ELIXIR node, hos ed a he CSC–IT Cen e o Science, Finland, and unded by he Academy o Finland (g an s 271642 and 263164), he Minis y o Educa ion and Cul u e, Finland. VS was suppo ed by he Finnish Academy (g an numbe 139635). Sample collec ion and geno yping in he Finnish Twin Coho has been suppo ed by he Wellcome T us Sange Ins i u e, ENGAGE–Eu opean Ne wo k o Gene ic and Genomic Epidemiology, FP7-HEALTH-F4-2007 (201413), he Na ional Ins i u e o Alcohol Abuse and Alcoholism (g an s AA-12502 and AA-00145 o Richa d J Rose, and K02AA018755 o Danielle M Dick) and he Academy o Finland (100499, 205585, 265240 and 263278 o Jaakko Kap io (JK)). The wo k o he Colon Cance Table 1. MR esul s showing he ela ionship be ween CRC isk, and adul BMI and adul WHR in he me a-analysis wi h published da a BMI WHR Sex OR (95% CI) P alue OR (95% CI) P alue All 1.26 (1.10–1.44) 7.73 10 4 1.40 (1.14–1.72) 1.22 10 3 Male 1.12 (0.92–1.38) 0.262 1.63 (1.20–2.22) 1.98 10 3 Female 1.43 (1.17–1.74) 4.25 10 4 1.22 (0.91–1.63) 0.178 Abb e ia ions: MR ¼Mendelian andomisa ion; CRC ¼colo ec al cance ; BMI ¼body mass index; WHR ¼wais -hip a io; OR ¼odds a io; CI ¼con idence in e al. Mendelian andomisa ion o adiposi y and CRC BRITISH JOURNAL OF CANCER www.bjcance .com | DOI:10.1038/bjc.2016.188 269 Family Regis y (CCFR) was suppo ed by he Na ional Cance Ins i u e (UM1 CA167551), Na ional Ins i u es o Heal h and h ough coope a i e ag eemen s wi h he ollowing CCFR cen es: Aus alasian Colo ec al Cance Family Regis y (U01 CA074778, U01/U24 CA097735), USC Conso ium Colo ec al Cance Family Regis y (U01/U24 CA074799), Mayo Clinic Coope a i e Familial Regis y o Colon Cance S udies (U01/U24 CA074800), On a io Familial Colo ec al Cance Regis y (U01/U24 CA074783), Sea le Colo ec al Cance Family Regis y (U01/U24 CA074794) and Uni e si y o Hawaii Colo ec al Cance Family Regis y (U01/U24 CA074806). The CCFR Illumina GWAS was suppo ed by unding om he Na ional Cance Ins i u e, Na ional Ins i u es o Heal h (U01 CA122839 and R01 CA143237 o GC). Sea le CCFR esea ch was also suppo ed by he Cance Su eillance Sys em o he F ed Hu chinson Cance Resea ch Cen e –Con ol Nos. N01-CN-67009 (1996–2003), N01-PC-35142 (2003–2010) and he Su eillance, Epidemiology and End Resul s (SEER) P og am o he Na ional Cance Ins i u e (Con ac No. HHSN2612013000121, 2010–2017) wi h addi ional suppo om he F ed Hu chinson Cance Resea ch Cen e . The collec ion o cance incidence da a o he S a e o Hawaii used in his s udy was suppo ed by he Hawai‘i Depa men o Heal h as pa o he s a ewide cance epo ing p og am manda ed by Hawai‘i Re ised S a u es; he Na ional Cance Ins i u e’s Su eillance, Epidemiology and End Resul s P og am (SEER) awa ded o he Uni e si y o Hawaii (Con ol Nos. N01-PC-67001 1996–2003, N01-PC-35137 2003–10, Con ac Nos. HHSN26120100037C 2010–13, HHSN261201300009I 2010-). The collec ion o cance incidence da a used in his s udy was suppo ed by he Cali o nia Depa men o Public Heal h as pa o he s a ewide cance epo ing p og am manda ed by Cali o nia Heal h and Sa e y Code Sec ion 103885; he Na ional Cance Ins i u e’s Su eillance, Epidemiology and End Resul s P og am— con ac HHSN261201000035C—awa ded o he Uni e si y o Sou he n Cali o nia, and—con ac HHSN261201000034C— awa ded o he Public Heal h Ins i u e; and he Cen e s o Disease Con ol and P e en ion’s Na ional P og am o Cance Regis ies, unde ag eemen —U58DP003862-01—awa ded o he Cali o nia Depa men o Public Heal h. CONFLICT OF INTEREST The au ho s decla e no con lic o in e es . DISCLAIMER The ideas and opinions exp essed he ein a e hose o he au ho (s) and endo semen by he S a e o Hawai‘i, Depa men o Heal h, he Na ional Cance Ins i u e, SEER P og am, he S a e o Cali o nia, Depa men o Public Heal h he Na ional Cance Ins i u e, and he Cen e s o Disease Con ol and P e en ion o hei Con ac o s and Subcon ac o s, is no in ended no should be in e ed. The con en o his manusc ip does no necessa ily e lec he iews o policies o he Na ional Cance Ins i u e o any o he collabo a ing cen e s in he CCFR, no does men ion o ade names, comme cial p oduc s o o ganisa ions imply endo semen by he US Go e nmen o he CCFR. REFERENCES Aleman JO, Eusebi LH, Riccia diello L, Pa ida K, Sanyal AJ, Hol PR (2014) Mechanisms o obesi y-induced gas oin es inal neoplasia. Gas oen e ology 146(2): 357–373. Ba dou M, Ba kun AN, Ma el M (2013) Obesi y and colo ec al cance . Gu 62(6): 933–947. Bianchini F, Kaaks R, Vainio H (2002) O e weigh , obesi y, and cance isk. Lance Oncol 3(9): 565–574. Bowden J, Da ey Smi h G, Bu gess S (2015) Mendelian andomiza ion wi h in alid ins umen s: e ec es ima ion and bias de ec ion h ough Egge eg ession. In J Epidemiol 44(2): 512–525. Bu gess S, Sco RA, Timpson NJ, Da ey Smi h G, Thompson SG. Conso ium E-I (2015) Using published da a in Mendelian andomiza ion: a bluep in o e icien iden i ica ion o causal isk ac o s. Eu J Epidemiol 30(7): 543–552. Ca pen e CL, Ross RK, Paganini-Hill A, Be ns ein L (2003) E ec o amily his o y, obesi y and exe cise on b eas cance isk among pos menopausal women. In J Cance 106(1): 96–102. Fo man D, B ay F, B ews e DH, Gombe Mbalawa C, Kohle B, Pin ˜e os M, S elia o a-Fouche E, Swamina han R, Fe lay J (2014) Cance incidence in i e con inen s IARC Scien i ic publica ions X(164): 917. F ayling TM, Timpson NJ, Weedon MN, Zeggini E, F ea hy RM, Lindg en CM, Pe y JR, Ellio KS, Lango H, Rayne NW, Shields B, Ha ies LW, Ba e JC, Ella d S, G o es CJ, Knigh B, Pa ch AM, Ness AR, Eb ahim S, Lawlo DA, Ring SM, Ben-Shlomo Y, Ja elin MR, So io U, Benne AJ, Melze D, Fe ucci L, Loos RJ, Ba oso I, Wa eham NJ, Ka pe F, Owen KR, Ca don LR, Walke M, Hi man GA, Palme CN, Doney AS, Mo is AD, Smi h GD, Ha e sley AT, McCa hy MI (2007) A common a ian in he FTO gene is associa ed wi h body mass index and p edisposes o childhood and adul obesi y. Science 316(5826): 889–894. Hagga FA, Boushey RP (2009) Colo ec al cance epidemiology: incidence, mo ali y, su i al, and isk ac o s. Clin Colon Rec al Su g 22(4): 191–197. Ha iss DJ, A kinson G, Geo ge K, Cable NT, Reilly T, Haboubi N, Zwahlen M, Egge M, Renehan AG. g oup CC (2009) Li es yle ac o s and colo ec al cance isk (1): sys ema ic e iew and me a-analysis o associa ions wi h body mass index. Colo ec al Dis 11(6): 547–563. Heid IM, Jackson AU, Randall JC, Winkle TW, Qi L, S ein ho sdo i V, Tho lei sson G, Zillikens MC, Spelio es EK, Magi R, Wo kalemahu T, Whi e CC, Boua ia-Naji N, Ha is TB, Be nd SI, Ingelsson E, Wille CJ, Weedon MN, Luan J, Vedan am S, Esko T, Kilpelainen TO, Ku alik Z, Li S, Monda KL, Dixon AL, Holmes CC, Kaplan LM, Liang L, Min JL, Mo a MF, Molony C, Nicholson G, Schad EE, Zonde an KT, Fei osa MF, Fe ei a T, Lango Allen H, Weyan RJ, Wheele E, Wood AR, Magic, Es ada K, Godda d ME, Le e G, Mangino M, Nyhol DR, Pu cell S, Smi h AV, Vissche PM, Yang J, McCa oll SA, Nemesh J, Voigh BF, Abshe D, Amin N, Aspelund T, Coin L, Glaze NL, Haywa d C, Hea d-Cos a NL, Ho enga JJ, Johansson A, Johnson T, Kaakinen M, Kapu K, Ke ka S, Knowles JW, K a P, K aja AT, Lamina C, Lei zmann MF, McKnigh B, Mo is AP, Ong KK, Pe y JR, Pe e s MJ, Polasek O, P okopenko I, Rayne NW, Ripa i S, Ri adenei a F, Robe son NR, Sanna S, So io U, Su akka I, Teume A, an Winge den S, Vi a V, Zhao JH, Ca alcan i-P oenca C, Chines PS, Fishe E, Kulze JR, Lecoeu C,Na isuN,Sandhol C,Sco LJ,Silande K,S a kK,TammesooML, Teslo ich TM, Timpson NJ, Wa anabe RM, Welch R, Chasman DI, Coope MN, Jansson JO, Ke unen J, Law ence RW, Pellikka N, Pe ola M, Vandenpu L, Ala e e H, Almg en P, A wood LD, Benne AJ, Bi a R, Bonnycas le LL, Bo ns ein SR, Buchanan TA, Campbell H, Day IN, Dei M, Do M, Ellio P, E dos MR, E iksson JG, F eime NB, Fu M, Gage S, Geus EJ, Gjesing AP, G alle H, G assle J, G o es CJ, Guiducci C, Ha ikainen AL, Hassanali N, Ha ulinna AS, He zig KH, Hicks AA, Hui J, Igl W, Jousilah i P, Jula A, Kajan ie E, Kinnunen L, Kolcic I, Koskinen S, Ko acs P, K oeme HK, K zelj V, Kuusis o J, K aloy K, Lai inen J, Lan ie i O, La h op GM, Lokki ML, Luben RN, Ludwig B, McA dle WL, McCa hy A, Mo ken MA, Nelis M, Ne ille MJ, Pa e G, Pa ke AN, Peden JF, Pichle I, Pie ilainen KH, Pla ou CG, Pou a A, Ridde s ale M, Samani NJ, Sa amies J, Sinisalo J, Smi JH, S awb idge RJ, S ingham HM, Swi AJ, Tede -La ing M, Thomson B, Usala G, an Meu s JB, an Ommen GJ, Va in V, Volpa o CB, Wallascho ski H, Wal e s GB, Widen E, Wild SH, Willemsen G, Wi e DR, Zgaga L, Zi ing P, Beilby JP, James AL, Kahonen M, Leh imaki T, Nieminen MS, Ohlsson C, Palme LJ, Rai aka i O, Ridke PM, S um oll M, Tonjes A, Viika i J, Balkau B, Ben-Shlomo Y, Be gman RN, Boeing H, Smi h GD, Eb ahim S, F oguel P, Hansen T, Hengs enbe g C, H eem K, Isomaa B, Jo gensen T, Ka pe F, Khaw KT, Laakso M, Lawlo DA, Ma e M, Mei inge T, Me spalu A, Mid hjell K, Pede sen O, Salomaa V, Schwa z PE, Tuomi T, Tuomileh o J, Valle TT, Wa eham NJ, A nold AM, Beckmann JS, Be gmann S, Boe winkle E, Boomsma DI, Caul ield MJ, Collins FS, Ei iksdo i G, Gudnason V, Gyllens en U, Hams en A, Ha e sley AT, Ho man A, Hu FB, Illig T, I iba en C, Ja elin MR, Kao WH, Kap io J, Laune LJ, BRITISH JOURNAL OF CANCER Mendelian andomisa ion o adiposi y and CRC 270 www.bjcance .com | DOI:10.1038/bjc.2016.188 Mun oe PB, Oos a B, Penninx BW, P ams alle PP, Psa y BM, Que e mous T, Rissanen A, Rudan I, Shuldine AR, So anzo N, Spec o TD, Sy anen AC, Uda M, Ui e linden A, Volzke H, Vollenweide P, Wilson JF, Wi eman JC, W igh AF, Abecasis GR, Boehnke M, Bo ecki IB, Deloukas P, F ayling TM, G oop LC, Ha i unians T, Hun e DJ, Kaplan RC, No h KE, O’Connell JR, Pel onen L, Schlessinge D, S achan DP, Hi schho n JN, Assimes TL, Wichmann HE, Tho s einsdo i U, an Duijn CM, S e ansson K, Cupples LA, Loos RJ, Ba oso I, McCa hy MI, Fox CS, Mohlke KL, Lindg en CM (2010) Me a-analysis iden i ies 13 new loci associa ed wi h wais -hip a io and e eals sexual dimo phism in he gene ic basis o a dis ibu ion. Na Gene 42(11): 949–960. Ho ikoshi M, Yaghoo ka H, Mook-Kanamo i DO, So io U, Taal HR, Hennig BJ, B ad ield JP, S Pou cain B, E ans DM, Cha oen P, Kaakinen M, Cousmine DL, Leh imaki T, K eine -Molle E, Wa ing on NM, Bus aman e M, Feens a B, Be y DJ, Thie ing E, P ab T, Ba on SJ, Shields BM, Ke kho M, an Leeuwen EM, Ful o d AJ, Ku alik Z, Zhao JH, den Hoed M, Mahajan A, Lindi V, Goh LK, Ho enga JJ, Wu Y, Rai aka i OT, Ha de MN, Mei haeghe A, N alla I, Salem RM, Jameson KA, Zhou K, Monies DM, Lagou V, Ki in M, Heikkinen J, Adai LS, Alku aya FS, Al-Odaib A, Amouyel P, Ande sson EA, Benne AJ, Blakemo e AI, Bux on JL, Dallonge ille J, Das S, de Geus EJ, Es i ill X, Flexede C, F oguel P, Gelle F, God ey KM, Go and F, G o es CJ, Hansen T, Hi schho n JN, Ho man A, Hollegaa d MV, Hougaa d DM, Hypponen E, Inskip HM, Isaacs A, Jo gensen T, Kanaka-Gan enbein C, Kemp JP, Kiess W, Kilpelainen TO, Klopp N, Knigh BA, Kuzawa CW, McMahon G, Newnham JP, Niinikoski H, Oos a BA, Pede sen L, Pos ma DS, Ring SM, Ri adenei a F, Robe son NR, Sebe S, Simell O, Slowinski T, Tiesle CM, Tonjes A, Vaag A, Viika i JS, Vink JM, Vissing NH, Wa eham NJ, Willemsen G, Wi e DR, Zhang H, Zhao J. Me a-Analyses o G, Insulin- ela ed ai s C, Wilson JF, S um oll M, P en ice AM, Meye BF, Pea son ER, Bo eham CA, Coope C, Gillman MW, Dedoussis GV, Mo eno LA, Pede sen O, Saa inen M, Mohlke KL, Boomsma DI, Saw SM, Lakka TA, Ko ne A, Loos RJ, Ong KK, Vollenweide P, an Duijn CM, Koppelman GH, Ha e sley AT, Holloway JW, Hoche B, Hein ich J, Powe C, Melbye M, Guxens M, Pennell CE, Bonnelykke K, Bisgaa d H, E iksson JG, Widen E, Hakona son H, Ui e linden AG, Pou a A, Lawlo DA, Smi h GD, F ayling TM, McCa hy MI, G an SF, Jaddoe VW, Ja elin MR, Timpson NJ, P okopenko I, F ea hy RM. Ea ly G ow h Gene ics C (2013) New loci associa ed wi h bi h weigh iden i y gene ic links be ween in au e ine g ow h and adul heigh and me abolism. Na Gene 45(1): 76–82. Kan P, Hull MA (2011) Excess body weigh and obesi y– he link wi h gas oin es inal and hepa obilia y cance . Na Re Gas oen e ol Hepa ol 8(4): 224–238. Lawlo DA, Ha bo d RM, S e ne JA, Timpson N, Da ey Smi h G (2008) Mendelian andomiza ion: using genes as ins umen s o making causal in e ences in epidemiology. S a Med 27(8): 1133–1163. Locke AE, Kahali B, Be nd SI, Jus ice AE, Pe s TH, Day FR, Powell C, Vedan am S, Buchko ich ML, Yang J, C o eau-Chonka DC, Esko T, Fall T, Fe ei a T, Gus a sson S, Ku alik Z, Luan J, Magi R, Randall JC, Winkle TW, Wood AR, Wo kalemahu T, Faul JD, Smi h JA, Hua Zhao J, Zhao W, Chen J, Feh mann R, Hedman AK, Ka jalainen J, Schmid EM, Abshe D, Amin N, Ande son D, BeekmanM,Bol onJL,B agg-G eshamJL, Buyske S, Demi kan A, Deng G, Eh e GB, Feens a B, Fei osa MF, Fische K, Goel A, Gong J, Jackson AU, Kanoni S, Klebe ME, K is iansson K, Lim U, Lo ay V, Mangino M, Ma eo Leach I, Medina-Gomez C, Medland SE, Nalls MA, Palme CD, Pasko D, Pechli anis S, Pe e s MJ, P okopenko I, Shungin D, S ancako a A, S awb idge RJ, Ju Sung Y, Tanaka T, Teume A, T ompe S, an de Laan SW, an Se en J, Van Vlie -Os ap chouk JV, Wang Z, Yengo L, Zhang W, Isaacs A, Alb ech E, A nlo J, A sco GM, A wood AP, Bandinelli S, Ba e A, Bas IN, Bellis C, Benne AJ, Be ne C, Blagie a R, Bluhe M, Boh inge S, Bonnycas le LL, Bo che Y, Boyd HA, B uinenbe g M, Caspe sen IH, Ida Chen YD, Cla ke R, Daw EW, de C aen AJ, Delgado G, Dimi iou M, Doney AS, Eklund N, Es ada K, Eu y E, Folke sen L, F ase RM, Ga cia ME, Gelle F, Gied ai is V, Gigan e B, Go AS, Golay A, Goodall AH, Go don SD, Go ski M, G abe HJ, G alle H, G amme TB, G assle J, G onbe g H, G o es CJ, Gus o G, Haessle J, Hall P, Halle T, Hallmans G, Ha man CA, Hassinen M, Haywa d C, Hea d-Cos a NL, Helme Q, Hengs enbe g C, Holmen O, Ho enga JJ, James AL, Je JM, Johansson A, Jolley J, Juliusdo i T, Kinnunen L, Koenig W, Kosken uo M, K a ze W, Lai inen J, Lamina C, Leande K, Lee NR, Lich ne P, Lind L, Linds om J, Sin Lo K, Lobbens S, Lo bee R, Lu Y, Mach F, Magnusson PK, Mahajan A, McA dle WL, McLachlan S, Menni C, Me ge S, Mihailo E, Milani L, Moayye i A, Monda KL, Mo ken MA, Mulas A, Mulle G, Mulle -Nu asyid M, Musk AW, Naga aja R, No hen MM, Nol e IM, Pilz S, Rayne NW, Rens om F, Re ig R, Ried JS, Ripke S, Robe son NR, Rose LM, Sanna S, Scha nagl H, Schol ens S, Schumache FR, Sco WR, Seu e lein T, Shi J, Ve non Smi h A, Smolonska J, S an on AV, S ein ho sdo i V, S i ups K, S ingham HM, Sunds om J, Swe z MA, Swi AJ, Sy anen AC, Tan ST, Tayo BO, Tho and B, Tho lei sson G, Ty e JP, Uh HW, Vandenpu L, Ve huls FC, Ve meulen SH, Ve weij N, Vonk JM, Wai e LL, Wa en HR, Wa e wo h D, Weedon MN, Wilkens LR, Willenbo g C, Wilsgaa d T, Wojczynski MK, Wong A, W igh AF, Zhang Q. Li eLines Coho SB ennan EP, Choi M, Das ani Z, D ong AW, E iksson P, F anco-Ce eceda A, Gadin JR, Gha a i AG, Godda d ME, Handsake RE, Huang J, Ka pe F, Ka hi esan S, Keildson S, Ki yluk K, Kubo M, Lee JY, Liang L, Li on RP, Ma B, McCa oll SA, McKnigh AJ, Min JL, Mo a MF, Mon gome y GW, Mu abi o JM, Nicholson G, Nyhol DR, Okada Y, Pe y JR, Do ajoo R, Reinmaa E, Salem RM, Sandholm N, Sco RA, S olk L, Takahashi A, Tanaka T, Van’ Hoo FM, Vinkhuyzen AA, Wes a HJ, Zheng W, Zonde an KT. Conso ium AD, G oup A-BW, Conso ium CAD, Conso ium CK, Glgc, Icbp, In es iga o s M, Mu TC, Conso ium MI, Conso ium P, Rep oGen C, Conso ium G, In e na ional Endogene CHea h AC, A eile D, Bakke SJ, Beilby J, Be gman RN, Blange o J, Bo e P, Campbell H, Caul ield MJ, Cesana G, Chak a a i A, Chasman DI, Chines PS, Collins FS, C aw o d DC, Cupples LA, Cusi D, Danesh J, de Fai e U, den Ruij e HM, Dominiczak AF, E bel R, E dmann J, E iksson JG, Fa all M, Felix SB, Fe annini E, Fe ie es J, Fo d I, Fo ouhi NG, Fo es e T, F anco OH, Ganse oo RT, Gejman PV, Giege C, Go esman O, Gudnason V, Gyllens en U, Hall AS, Ha is TB, Ha e sley AT, Hicks AA, Hindo LA, Hingo ani AD, Ho man A, Homu h G, Ho ingh GK, Humph ies SE, Hun SC, Hypponen E, Illig T, Jacobs KB, Ja elin MR, Jockel KH, Johansen B, Jousilah i P, Jukema JW, Jula AM, Kap io J, Kas elein JJ, Keinanen-Kiukaanniemi SM, Kiemeney LA, Knek P, Koone JS, Koope be g C, Ko acs P, K aja AT, Kuma i M, Kuusis o J, Lakka TA, Langenbe g C, Le Ma chand L, Leh imaki T, Lyssenko V, Mannis o S, Ma e e A, Ma ise TC, McKenzie CA, McKnigh B, Moll FL, Mo is AD, Mo is AP, Mu ay JC, Nelis M, Ohlsson C, Oldehinkel AJ, Ong KK, Madden PA, Pas e kamp G, Peden JF, Pe e s A, Pos ma DS, P ams alle PP, P ice JF, Qi L, Rai aka i OT, Rankinen T, Rao DC, Rice TK, Ridke PM, Rioux JD, Ri chie MD, Rudan I, Salomaa V, Samani NJ, Sa amies J, Sa zynski MA, Schunke H, Schwa z PE, Se e P, Shuldine AR, Sinisalo J, S olk RP, S auch K, Tonjes A, T egoue DA, T emblay A, T emoli E, Vi amo J, Vohl MC, Volke U, Waebe G, Willemsen G, Wi eman JC, Zillikens MC, Adai LS, Amouyel P, Asselbe gs FW, Assimes TL, Bochud M, Boehm BO, Boe winkle E, Bo ns ein SR, Bo inge EP, Boucha d C, Cauchi S, Chambe s JC, Chanock SJ, Coope RS, de Bakke PI, Dedoussis G, Fe ucci L, F anks PW, F oguel P, G oop LC, Haiman CA, Hams en A, Hui J, Hun e DJ, H eem K, Kaplan RC, Ki imaki M, Kuh D, Laakso M, Liu Y, Ma in NG, Ma z W, Melbye M, Me spalu A, Moebus S, Mun oe PB, Njols ad I, Oos a BA, Palme CN, Pede sen NL, Pe ola M, Pe usse L, Pe e s U, Powe C, Que e mous T, Rau amaa R, Ri adenei a F, Saa is o TE, Saleheen D, Sa a N, Schad EE, Schlessinge D, Slagboom PE, Sniede H, Spec o TD, Tho s einsdo i U, S um oll M, Tuomileh o J, Ui e linden AG, Uusi upa M, an de Ha s P, Walke M, Wallascho ski H, Wa eham NJ, Wa kins H, Wei DR, Wichmann HE, Wilson JF, Zanen P, Bo ecki IB, Deloukas P, Fox CS, Heid IM, O’Connell JR, S achan DP, S e ansson K, an Duijn CM, Abecasis GR, F anke L, F ayling TM, McCa hy MI, Vissche PM, Sche ag A, Wille CJ, Boehnke M, Mohlke KL, Lindg en CM, Beckmann JS, Ba oso I, No h KE, Ingelsson E, Hi schho n JN, Loos RJ, Spelio es EK (2015) Gene ic s udies o body mass index yield new insigh s o obesi y biology. Na u e 518(7538): 197–206. Moghaddam AA, Woodwa d M, Huxley R (2007) Obesi y and isk o colo ec al cance : a me a-analysis o 31 s udies wi h 70,000 e en s. Cance Epidemiol Bioma ke s P e 16(12): 2533–2547. O lando G, Law PJ, Palin K, Tuupanen S, Gyl e A, Ha ¨nninen U, Cajuso T, Tanskanen T, Kondelin J, Kaasinen E, Sa in A-P, Kap io J, E iksson JG, Rissanen H, Knek P, Pukkala E, Jousilah i P, Salomaa V, Ripa i S, Palo ie A, Ja ¨ inen H, Renkonen-Sinisalo L, Lepis o ¨A, Bo ¨hm J, Meklin J-P, Al-Tassan NA, Palles C, Ma in L, Ba clay E, Tenesa A, Fa ing on S, Timo ee a MN, Meye BF, Wakil SM, Campbell H, Mendelian andomisa ion o adiposi y and CRC BRITISH JOURNAL OF CANCER www.bjcance .com | DOI:10.1038/bjc.2016.188 271 Smi h CG, Idziaszczyk S, Maughan TS, Kaplan R, Ke R, Ke D, Buchanan DD, Win AK, Hoppe J, Jenkins M, Lindo NM, Newcomb PA, Gallinge S, Con i D, Schumache F, Casey G, Taipale J, Cheadle JP, Dunlop MG, Tomlinson IP, Aal onen LA, Houls on RS (2016) Va ia ion a 2q35 (PNKD and TMBIM1) in luences colo ec al cance isk and iden i ies a pleio opic e ec wi h in lamma o y bowel disease. Hum Mol Gene . Pan SY, Johnson KC, Ugna AM, Wen SW, Mao Y. Canadian Cance Regis ies Epidemiology Resea ch G (2004) Associa ion o obesi y and cance isk in Canada. Am J Epidemiol 159(3): 259–268. Rohde K, Kelle M, Klos M, Schleini z D, Die ich A, Schon MR, Ga ne D, Lohmann T, D essle M, S um oll M, Ko acs P, Bluhe M, Bo che Y (2014) Adipose issue depo speci ic p omo e me hyla ion o TMEM18. J Mol Med 92(8): 881–888. Se dula MK, I e y D, Coa es RJ, F eedman DS, Williamson DF, Bye s T (1993) Do obese child en become obese adul s? A e iew o he li e a u e. P e en Med 22(2): 167–177. Smi h GD, Eb ahim S (2003) ‘Mendelian andomiza ion’: can gene ic epidemiology con ibu e o unde s anding en i onmen al de e minan s o disease? In J Epidemiol 32(1): 1–22. Th i AP, Gong J, Pe e s U, Chang-Claude J, Rudolph A, Sla e y ML, Chan AT, Locke AE, Kahali B, Jus ice AE, Pe s TH, Gallinge S, Hayes RB, Ba on JA, Caan BJ, Ogino S, Be nd SI, Chanock SJ, Casey G, Haile RW, Du M, Ha ison TA, Tho nquis M, Duggan DJ, Le Ma chand L, Lindo NM, Semina a D, Song M, Wu K, Thibodeau SN, Co e chio M, Win AK, Jenkins MA, Hoppe JL, Ul ich CM, Po e JD, Newcomb PA, Ho meis e M, B enne H, Whi e E, Hsu L, Campbell PT (2015) Mendelian andomiza ion s udy o body mass index and colo ec al cance isk. Cance Epidemiol Bioma ke s P e 24(7): 1024–1031. Wheele E, Huang N, Bochuko a EG, Keogh JM, Lindsay S, Ga g S, Henning E, Blackbu n H, Loos RJ, Wa eham NJ, O’Rahilly S, Hu les ME, Ba oso I, Fa ooqi IS (2013) Genome-wide SNP and CNV analysis iden i ies common and low- equency a ian s associa ed wi h se e e ea ly-onse obesi y. Na Gene 45(5): 513–517. Yamauchi M, Lochhead P, Mo ikawa T, Hu enhowe C, Chan AT, Gio annucci E, Fuchs C, Ogino S (2012) Colo ec al cance : a ale o wo sides o a con inuum? Gu 61(6): 794–797. Yang TO, Ree es GK, G een J, Be al V, Cai ns BJ. Million Women S udy C, Million Women S udy C (2014) Bi h weigh and adul cance incidence: la ge p ospec i e s udy and me a-analysis. Ann Oncol 25(9): 1836–1843. This wo k is licensed unde he C ea i e Commons A ibu ion 4.0 In e na ional License. To iew a copy o his license, isi h p://c ea i ecommons.o g/licenses/by/4.0/ Supplemen a y In o ma ion accompanies his pape on B i ish Jou nal o Cance websi e (h p://www.na u e.com/bjc) 1 Di ision o Gene ics and Epidemiology, The Ins i u e o Cance Resea ch, London SW7 3RP, UK; 2 Genome-Scale Biology Resea ch P og am, Resea ch P og ams Uni , Uni e si y o Helsinki 00014, Helsinki, Finland; 3 Depa men o Medical and Clinical Gene ics, Medicum, Uni e si y o Helsinki, Helsinki 00014, Finland; 4 Ins i u e o Molecula Medicine Finland, Uni e si y o Helsinki, Helsinki 00014, Finland; 5 Na ional Ins i u e o Heal h and Wel a e, Helsinki 00271, Finland; 6 Folkha ¨lsan Resea ch Cen e, Helsinki 00250, Finland; 7 Uni o Gene al P ac ice and P ima y Heal h Ca e, Uni e si y o Helsinki, Helsinki Uni e si y Hospi al, Helsinki 00014, Finland; 8 Finnish Cance Regis y, Ins i u e o S a is ical and Epidemiological Cance Resea ch, Helsinki 00130, Finland; 9 School o Heal h Sciences, Uni e si y o Tampe e, Tampe e 33014, Finland; 10 Depa men o Medicine, Analy ic and T ansla ional Gene ics Uni , Massachuse s Gene al Hospi al, Bos on, MA 02114, USA; 11 P og am in Medical and Popula ion Gene ics, The B oad Ins i u e o MIT and Ha a d, Camb idge, MA 02142, USA; 12 Depa men o Neu ology, Massachuse s Gene al Hospi al, Bos on, MA 02114, USA; 13 Depa men o Su ge y, Helsinki Uni e si y Cen al Hospi al, Hospi al Dis ic o Helsinki and Uusimaa, Helsinki 00029, Finland; 14 Depa men o Su ge y, Abdominal Cen e , Helsinki Uni e si y Hospi al Helsinki 00029, Finland; 15 Depa men o Pa hology, Cen al Finland Cen al Hospi al, Jy a ¨skyla ¨40620, Finland; 16 Depa men o Su ge y, Jy a ¨skyla ¨Cen al Hospi al, Uni e si y o Eas e n Finland, Jy a ¨skyla ¨40620, Finland; 17 Depa men o Gene ics, King Faisal Specialis Hospi al and Resea ch Cen e , Riyadh 12713, Saudi A abia; 18 Wellcome T us Cen e o Human Gene ics, NIHR Comp ehensi e Biomedical Resea ch Cen e, Ox o d OX3 7BN, UK; 19 Colon Cance Gene ics G oup, MRC Human Gene ics Uni , The Uni e si y o Edinbu gh, Wes e n Gene al Hospi al, Edinbu gh EH4 2XU, UK; 20 The Roslin Ins i u e, Uni e si y o Edinbu gh, Eas e Bush, Roslin, Edinbu gh EH25 9RG, UK; 21 Cen e o Popula ion Heal h Sciences, Uni e si y o Edinbu gh, Edinbu gh EH8 9AG, UK; 22 Ins i u e o Cance and Gene ics, School o Medicine, Ca di Uni e si y, Ca di CF14 4XN, UK; 23 CRUK/ MRC Ox o d Ins i u e o Radia ion Oncology, Uni e si y o Ox o d, Ox o d OX3 7DQ, UK; 24 MRC Clinical T ials Uni , A ia ion House, London WC2B 6NH, UK; 25 Depa men o Oncology, Ox o d Cance Cen e, Uni e si y o Ox o d, Chu chill Hospi al, Ox o d OX3 7LE, UK; 26 Nu ield Depa men o Clinical Labo a o y Sciences, Uni e si y o Ox o d, John Radcli e Hospi al, Ox o d OX3 9DU, UK; 27 Depa men o Pa hology, Colo ec al Oncogenomics G oup, Gene ic Epidemiology Labo a o y, The Uni e si y o Melbou ne, Melbou ne, VIC 3010, Aus alia; 28 Cen e o Epidemiology and Bios a is ics, The Uni e si y o Melbou ne, Melbou ne, VIC 3010, Aus alia; 29 Depa men o Heal h Sciences Resea ch, Mayo Clinic, Sco sdale, AZ 85259, USA; 30 Cance P e en ion P og am, F ed Hu chinson Cance Resea ch Cen e , Sea le, WA 98109, USA; 31 Lunen eld- Tanenbaum Resea ch Ins i u e, Moun Sinai Hospi al, To on o, ON M5G 1X5, Canada; 32 Depa men o P e en i e Medicine, Uni e si y o Sou he n Cali o nia, Los Angeles, CA 90033, USA and 33 Depa men o Biosciences and Nu i ion, SciLi e Cen e , Ka olinska Ins i u e , Solna SE 141 83, Sweden BRITISH JOURNAL OF CANCER Mendelian andomisa ion o adiposi y and CRC 272 www.bjcance .com | DOI:10.1038/bjc.2016.188