Endogenous Circulating Sex Hormone Concentrations and Colon Cancer Risk in Postmenopausal Women: A Prospective Study and Meta-Analysis

Endogenous Ci cula ing Sex Ho mone Concen a ions and Colon Cance Risk in Pos menopausal Women: A P ospec i e S udy and Me a-Analysis Nagisa Mo i , PhD, 1, * Pekka Keski-Rahkonen , PhD, 1 Aud ey Gicquiau, MSc, 1 Sabina Rinaldi , PhD, 1 Niki Dimou, PhD, 1 Sophia Ha lid , PhD, 2 Jus in Ha bs , MSc, 2 Be hany Van Guelpen, PhD, 2,3 Dag inn Aune , PhD, 4,5,6 Amanda J. C oss , PhD, 4 Kons an inos K. Tsilidis , PhD, 4,7 Gianluca Se e i , PhD, 8,9 Ma ina K asko , PhD, 8 Agne`s Fou nie , PhD, 8 Rudol Kaaks, PhD, 10 Ren ee Tu zanski Fo ne , PhD, 10 Ma hias B. Schulze , PhD, 11 Paula Jakszyn, PhD, 12 Ma ia-Jose S anchez, PhD, 13,14,15,16 Sand a M. Colo ado-Yoha , PhD, 17,18,19 E a A danaz , PhD, 15,20,21 Ru h T a is, PhD, 22 Eleano L. Wa s , PhD, 22 Gio anna Masala , PhD, 23 Vi o io K ogh , PhD, 24 Rosa io Tumino , PhD, 25 Ca lo a Sace do e , PhD, 26 Sal a o e Panico, PhD, 27 Bas Bueno-de-Mesqui a, PhD, 28 Inge To hild G am , PhD, 29 Ma i Waase h, PhD, 30 Ma c J. Gun e , PhD, 1 Neil Mu phy, PhD 1 1 Nu i ion and Me abolism B anch, In e na ional Agency o Resea ch on Cance , Lyon, F ance, 2 Depa men o Radia ion Sciences, Oncology, Umea˚ Uni e si y, Umea˚, Sweden, 3 Wallenbe g Cen e o Molecula Medicine, Umea˚ Uni e si y, Umea˚, Sweden, 4 Depa men o Epidemiology and Bios a is ics, Impe ial College London, No olk Place, London, UK, 5 Depa men o Nu i ion, Bjø knes Uni e si y College, Oslo, No way, 6 Depa men o Endoc inology, Mo bid Obesi y and P e en i e Medicine, Oslo Uni e si y Hospi al, Ulle a˚l, Oslo, No way, 7 Depa men o Hygiene and Epidemiology, Uni e si y o Ioannina School o Medicine, Ioannina, G eece, 8 Pa is-Saclay Uni e si y, UVSQ, Inse m, Gus a e Roussy, “Exposome and He edi y” eam, CESP, Villejui , F ance, 9 Depa men o S a is ics, Compu e Science, Applica ions “G. Pa en i,” Uni e si y o Flo ence, Flo ence, I aly, 10 Depa men o Cance Epidemiology, Ge man Cance Resea ch Cen e (DKFZ), Heidelbe g, Ge many, 11 Depa men o Molecula Epidemiology, Ge man Ins i u e o Human Nu i ion, Po sdam, Ge many, 12 Uni o Nu i ion and Cance , Cance Epidemiology Resea ch P og amme, Ca alan Ins i u e o Oncology (ICO-IDIBELL), Ba celona, Spain, 13 Escuela Andaluza de Salud P ublica (EASP), G anada, Spain, 14 Ins i u o de In es igaci on Biosani a ia ibs.GRANADA, G anada, Spain, 15 Cen o de In es igaci on Biom edica en Red de Epidemiolog ıa y Salud P ublica (CIBERESP), Mad id, Spain, 16 Depa men o P e en i e Medicine and Public Heal h, Uni e si y o G anada, G anada, Spain, 17 Depa men o Epidemiology, Mu cia Regional Heal h Council, IMIB-A ixaca, Mu cia, Spain, 18 CIBER Epidemiolog ıa y Salud P ublica (CIBERESP), Mad id Spain, 19 Resea ch G oup on Demog aphy and Heal h, Na ional Facul y o Public Heal h, Uni e si y o An ioquia, Medell ın, Colombia, 20 Na a a Public Heal h Ins i u e, Pamplona, Spain, 21 IdiSNA, Na a a Ins i u e o Heal h Resea ch, Pamplona, Spain, 22 Cance Epidemiology Uni , Nu ield Depa men o Popula ion Heal h, Uni e si y o Ox o d, Ox o d, UK, 23 Ins i u e o Cance Resea ch, P e en ion and Clinical Ne wo k— ISPRO, Flo ence, I aly, 24 Epidemiology and P e en ion Uni , Fondazione IRCCS Is i u o Nazionale dei Tumo i di Milano, Milan, I aly, 25 Cance Regis y and His opa hology Depa men , P o incial Heal h Au ho i y (ASP 7), Ragusa, I aly, 26 Uni o Cance Epidemiology, Piedmon Child en Cance Regis y, Ci  a della Salu e e della Scienza Uni e si y-Hospi al and Cen e o Cance P e en ion (CPO), Tu in, I aly, 27 Dipa imen o di Medicina Clinica e Chi u gia, Fede ico II Uni e si y, Naples, I aly, 28 Cen e o Nu i ion, P e en ion and Heal h Se ices, Na ional Ins i u e o Public Heal h and he En i onmen , Bil ho en, The Ne he lands, 29 Facul y o Heal h Sciences, Depa men o Communi y Medicine, UiT The A c ic Uni e si y o No way, T omsø, No way and 30 Depa men o Pha macy, The Facul y o Heal h Sciences, UiT The A c ic Uni e si y o No way, T omsø, No way *Co espondence o: Nagisa Mo i, PhD, Nu i ion and Me abolism B anch, In e na ional Agency o Resea ch on Cance , 150 Cou s Albe Thomas, 69372 Lyon, Cedex 08, F ance (e-mail: [email p o ec ed]). Abs ac Backg ound: Obse a ional s udies ha e consis en ly epo ed ha pos menopausal ho mone he apy use is associa ed wi h lowe colon cance isk, bu epidemiologic s udies examining he associa ions be ween ci cula ing concen a ions o endogenous es ogens and colo ec al cance ha e epo ed inconsis en esul s. Me hods: We in es iga ed he associa ions be ween ci cula ing concen a ions o es one, es adiol, ee es adiol, es os e one, ee es os e one, and os enedione, dehyd oepiand os e one (DHEA), p oges e one, and sex ho mone–binding globulin (SHBG) wi h colon cance isk in a nes ed case-con ol s udy o 1028 pos menopausal Eu opean women (512 colon cance cases, 516 ma ched con ols) who we e non- cu en use s o exogenous ho mones a blood collec ion. Mul i a iable condi ional logis ic eg ession models we e used o compu e odds a ios and 95% con idence in e als o e alua e he associa ion be ween ci cula ing sex ho mones and colon Recei ed: 18 May 2021; Re ised: 5 Augus 2021; Accep ed: 27 Augus 2021 ©The Au ho (s) 2021. Published by Ox o d Uni e si y P ess. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion-NonComme cial-NoDe i s licence (h ps://c ea i ecom- mons.o g/licenses/by-nc-nd/4.0/), which pe mi s non-comme cial ep oduc ion and dis ibu ion o he wo k, in any medium, p o ided he o iginal wo k is no al e ed o ans o med in any way, and ha he wo k is p ope ly ci ed. Fo comme cial e-use, please con ac [email p o ec ed] 1o 10 JNCI Cance Spec um (2021) 5(6): pkab084 doi: 10.1093/jncics/pkab084 Fi s published online 28 Sep embe 2021 A icle Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 cance isk. We also conduc ed a dose- esponse me a-analysis o p ospec i e s udies o ci cula ing es one and es adiol wi h colo ec al, colon, and ec al cance isk in pos menopausal women. All s a is ical es s we e 2-sided. Resul s: In he mul i a iable model, a nons a is ically signi ican ly posi i e ela ionship was ound be ween ci cula ing es one and colon cance isk (odds a io pe log 2 1-uni inc emen ¼1.17 [95% con idence in e al¼1.00 o 1.38]; odds a io qua ile4-qua ile1 ¼1.33 [95% con idence in e al¼0.89 o 1.97], P end ¼.20). Ci cula ing concen a ions o es adiol, ee es adiol, es os e one, ee es os e one, and os enedione, DHEA, p oges e one, and SHBG we e no associa ed wi h colon cance isk. In he dose- esponse me a-analysis, no clea e idence o associa ions we e ound be ween ci cula ing es adiol and es one concen a- ions wi h colo ec al, colon, and ec al cance isk. Conclusion: Ou obse a ional and me a-analysis esul s do no suppo an associa ion be ween ci cula ing concen a ions o endogenous sex ho mones and colon o ec al cance in pos meno- pausal women. Colo ec al cance is he hi d-mos common cance globally, wi h a lowe incidence gene ally ound o women han o men (1). I has been hypo hesized ha he sex dispa i y in incidence may be explained by highe es ogen concen a ions in women, con e ing a p o ec i e ole agains colon cance de elopmen (2). Consis en wi h his hypo hesis, mul iple obse a ional s udies and a clinical ial ha e ound ha he use o pos menopausal ho mone he apy (HT) was associa ed wi h lowe colo ec al cance isk in women (3-7). Epidemiologic da a on he associa ion o endogenous es o- gens and o he sex ho mones wi h colo ec al umo igenesis a e ela i ely limi ed. Ini ial analyses o endogenous ci cula ing sex ho mone concen a ions and colo ec al cance isk in pos meno- pausal women did no suppo an an i umo igenic e ec o es ogens in he colo ec um (8-11), bu in a mo e ecen case- con ol s udy nes ed wi hin he Women’s Heal h Ini ia i e Clinical T ial (WHI-CT), in e se associa ions we e epo ed be- ween endogenous es ogens and colo ec al and colon cance isk bu no ec al cance , while a posi i e ela ionship be ween sex ho mone–binding globulin (SHBG) and colo ec al cance isk was obse ed (12). Addi ional s udies a e needed o p o ide mo e cla i y on he ole o es ogens in colo ec al umo igenesis. Tes os e one is a biologically po en and ogen and he main sou ce o es adiol in women a e menopause (13). The ole o es os e one in ela ion o colo ec al cance in pos menopausal women is unce ain, bu a ecen nes ed case-con ol s udy con- duc ed among pos menopausal Japanese women epo ed a posi i e associa ion be ween endogenous es os e one concen- a ions and colo ec al cance isk (14). Fu he p ospec i e s udies a e wa an ed o examine he ole o es os e one and o he and ogens, dehyd oepiand os e one (DHEA), and and o- s enedione in colo ec al cance de elopmen . To p o ide mo e conclusi e e idence o he associa ion be- ween endogenous concen a ions o ci cula ing sex ho mones and colon cance , we conduc ed a nes ed case-con ol s udy wi hin he Eu opean P ospec i e In es iga ion in o Cance and Nu i ion (EPIC) coho and he No he n Sweden Heal h and Disease S udy (NSHDS) coho in which ci cula ing concen a- ions o es adiol, es one, es os e one, and os enedione, DHEA, p oges e one, and SHBG we e measu ed. In addi ion, we con- duc ed a me a-analysis combining esul s om he cu en s udy wi h hose om p e iously published p ospec i e s udies (8- 12,14) o examine he o e all e idence linking endogenous es a- diol and es one wi h colo ec al, colon, and ec al cance isk. Me hods S udy Popula ion and Collec ion o Blood Samples and Da a EPIC is an ongoing mul icen e p ospec i e coho o 521 330 pa icipan s who we e ec ui ed be ween 1992 and 2000, p edominan ly om he gene al popula ion o 10 Eu opean coun ies (Denma k, F ance, Ge many, G eece, I aly, he Ne he lands, No way, Spain, Sweden, and he Uni ed Kingdom) (15-17). Blood samples we e collec ed a he ime o ec ui men by s anda dized p ocedu es (16,17) and s o ed a he In e na ional Agency o Resea ch on Cance (IARC) (–196C, liq- uid ni ogen) excep o Denma k (–150C, ni ogen apo ) and Sweden (–80C, eeze s). All pa icipan s comple ed li es yle ques ionnai es a ec ui men , and mos o he pa icipan s had an h opome ic measu emen s and comple ed a alida ed ood equency ques ionnai e. All pa icipan s p o ided w i en in- o med consen a ec ui men . E hical app o al o he s udy was ob ained om he e iew boa ds o IARC and om local pa icipa ing cen e s. NSHDS is an ongoing popula ion-based coho o 135 000 pa icipan s ha began in 1985. I consis s o 3 subcoho s: he V€ as e bo en In e en ion P og amme, he mammog aphy sc eening coho , and he No he n Sweden MONI o ing o ends and de e minan s in CA dio ascula disease (MONICA) s udy (18,19). App oxima ely 80% o he V€ as e bo en In e en ion P og amme and MONICA coho pa icipan s do- na ed hei blood a e o e nigh as ing. In he mammog aphy sc eening coho , he ime since he las meal was eco ded. All blood samples we e s o ed in eeze s a –80C. A ec ui men , all pa icipan s unde wen a heal h examina ion (including measu emen o heigh and weigh ) and comple ed a alida ed ood equency ques ionnai e and li es yle ques ionnai e. The s udy was app o ed by he Resea ch E hics Commi ee o Umea˚ Uni e si y Hospi al and he Regional E hics Commi ee in Uppsala (No. 2013-124). Follow-up o Cance Incidence In o ma ion abou cance incidence was e ie ed om local cance egis ies, excep o F ance and Ge many, whe e inci- den cases we e iden i ied h ough a combina ion o heal h in- su ance eco ds, cance and pa hology egis ies, and ac i e ollow-up o pa icipan s (20,21). G eece we e excluded om he cu en analysis because o an ongoing adminis a i e issue. Colon cance cases we e coded acco ding o he In e na ional Classi ica ion o Diseases o Oncology, Thi d Edi ion (C18-C20) (22). We included colon cance cases wi hin he p oximal (C18.0- 18.5), dis al (C18.6-C18.7), o e lapping (C18.8), and unspeci ied egions (C18.9). Selec ion o Cases and Con ols Because o unding cons ain s and o e iciency, we used a nes ed case-con ol design. As ou ocus was on endogenous ci cula ing sex ho mone concen a ions, ou s udy was limi ed 2o 10 | JNCI Cance Spec um, 2021, Vol. 5, No. 6 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 o pos menopausal women who we e no aking menopausal HT when blood samples we e collec ed. Addi ionally, women who sel - epo ed diabe es a baseline o hose wi h unknown diabe ic s a us we e excluded [because diabe es a ec s he con- cen a ions o sex ho mones (23-25)]. Fo he selec ion o con- ols among EPIC pa icipan s, we sampled om all pa icipan s who we e li ing, noncu en use s o exogenous ho mones; pos menopausal; and ee o cance (excep nonmelanoma skin cance ) a he ime o diagnosis o he cases, using incidence densi y sampling and ma ched by age, s udy cen e , ollow-up ime since blood collec ion, ime o day a blood collec ion, and as ing s a us. Fo NSHDS pa icipan s, we ma ched by s udy cen e , ollow-up ime, age, da e o blood collec ion, and as ing s a us. We iden i ied 557 inciden colon cance cases and 564 ma ched con ols. O hese, we excluded pa icipan s wi h missing es adiol concen a ion (n ¼1) and hose in unma ched case-con ol se s (n ¼92). A e hese exclusions, he cu en s udy included 512 colon cance cases and 516 ma ched con- ols. O hese, 109 pa icipan s we e also pa o he NSHDS co- ho bu we e ega ded as EPIC pa icipan s in he cu en s udy. The 26 cases and 26 ma ched con ols we e hose pa icipan s unique o NSHDS and no he wide EPIC s udy. Labo a o y Me hods and Assessmen o Sex Ho mone Concen a ions and SHBG Plasma sex ho mones and SHBG concen a ions we e measu ed a IARC (Lyon, F ance) using a alida ed analy ical me hod adap ed om p e ious publica ions (26,27). Full de ails a e in- cluded in he Supplemen a y Me hods (a ailable online). Sex ho mones we e assayed using in a liquid ch oma og aphy– mass spec ome y sys em consis ing o a ul a-high-pe o - mance liquid ch oma og aph (Agilen 1290, Agilen , San a Cla a, CA) and a QTRAP 5500 mass spec ome e (SCIEX, F amingham, MA). Solid-phase “sandwich” enzyme-linked immunoassay (DRG In e na ional, Sp ing ield, NJ) was used o he measu e- men o SHBG concen a ions. Lowe limi s o quan i ica ion (LLOQ) o each sex ho mone was 7.5 pg/mL o and os enedi- one, 125 pg/mL o DHEA, 1.25 pg/mL o es adiol, 1.25 pg/mL o es one, 7.5 pg/mL o p oges e one, 7.5 pg/mL o es os e one, and 4 nmol/L o SHBG. Th ee quali y con ol samples a di e - en concen a ion le els we e measu ed in duplica e in each ba ch o analyses. In aba ch coe icien s o a ia ion in sex ho - mone and SHBG concen a ions anged om 1.4% o 8%. In e ba ch coe icien s o a ia ions we e less han 10% o all analy es. Plasma concen a ion o ee es adiol was calcula ed using a alida ed algo i hm (28), aking in o conside a ion measu ed es adiol and SHBG concen a ions and an assumed cons an o albumin. F ee es os e one was also compu ed om p e i- ously alida ed mass-ac ion equa ion using absolu e concen a- ions o es os e one and an assumed albumin cons an o 43 g/ L(29,30). We also calcula ed he es adiol- o- es os e one a io (by di iding he es adiol concen a ion by he es os e one con- cen a ion) because a highe a io indica es g ea e p oduc ion o es adiol om a oma ase con e sion. S a is ical Analysis We impu ed alues o be hal he LLOQ o hose pa icipan s wi h a lowe LLOQ alue (4.4% o es adiol and 0.1% o DEHA). Pea son co ela ion coe icien s (adjus ed o age a blood col- lec ion and ba ch) be ween ci cula ing concen a ions o log 2 - ans o med sex ho mones and SHBG and body mass index (BMI) we e calcula ed o con ol pa icipan s. Pa icipan s in he con ol g oup we e di ided in o ei he e iles o qua iles based on sex ho mone concen a ions. S a is ical es s o ends in he p esen analyses we e pe o med using he o dinal e ile o qua ile en e ed in o he models as con inuous a ia- bles. We also conduc ed con inuous analyses wi h each log 2 - ans o med sex ho mone. Mul i a iable condi ional eg ession models, s a i ied by case–con ol se , we e used o examine he associa ion be ween ci cula ing sex ho mone concen a ions and colon cance isk. The mul i a iable models we e adjus ed o BMI, smoking s a us, physical ac i i y, e e HT use, and alco- hol consump ion. In u he analyses, he es one, es adiol, and es os e one models we e addi ionally adjus ed o SHBG and ice e sa. Fu he adjus men o die a y in akes o o al ene gy, die a y ibe , and ed and p ocessed mea esul ed in simila esul s, so hese a iables we e no included in he inal mul i a iable models. False-disco e y a e co ec ion was pe - o med using he Benjamini-Hochbe g me hod o he main analysis (31). We also examined he sex ho mone and colon cance associ- a ions acco ding o subg oups o BMI (<25 kg/m 2 and 25 kg/m 2 ) and ollow-up ime (below o abo e median ollow-up in yea s) and compu ed he P alue o in e ac ion wi h he addi ion o an in e ac ion e m in he model by he a o emen ioned ca ego- ies. We used a likelihood a io es o assess he di e ence be- ween he models wi h and wi hou he in e ac ion e m. Analyses o p oximal colon cance and dis al colon cance we e also conduc ed. In u he analyses, o assess he po en ial in luence o p eclinical disease on he esul s, cases diagnosed wi hin he i s 2 yea s o ollow-up we e excluded. In an addi- ional sensi i i y analysis, we used a mul iple-impu a ion p o- cedu e (SAS command: PROC MI [SAS Ins i u e Inc, Ca y, NC]) o impu e alues below he LLOQ (es adiol, DHEA, es adiol- o- es os e one a io, and ee es adiol models only) (32,33). All s a is ical analyses we e pe o med using SAS so wa e, e sion 9.4. All s a is ical es s we e 2-sided. Me a-Analysis We pe o med a hand sea ch up o July 2020 on PubMed using he keywo ds (“colo ec al” OR “colo ec um” OR “colon” OR “ ec um”) and “cance ” and “sex ho mone.” We limi ed ou sea ch o s udies published in English ha p ospec i ely e alu- a ed he associa ion be ween ci cula ing es adiol and es one wi h colo ec al, colon, o ec al cance isk. Full de ails o he me a-analysis me hods a e desc ibed in he Supplemen a y Me hods (a ailable online). We calcula ed summa y ela i e isks (RRs) and 95% con i- dence in e als (CIs) o a 5 pg/mL inc emen in es adiol and a 10 pg/mL inc emen in es one using a andom e ec s model. The a e age o he na u al loga i hm o he ela i e isks was es ima ed, and he ela i e isk o each s udy was weigh ed us- ing andom e ec s weigh ing. The dose- esponse analysis desc ibed by G eenland and Longnecke (34) was used o compu e speci ic slopes (linea ends) and 95% con idence in e als om he na u al logs o he epo ed ela i e isks and con idence in e als ac oss ca e- go ies o es adiol and es one concen a ions. He e ogenei y in esul s ac oss s udies was also examined using Coch an Qand I 2 s a is ics. S a is ical analyses we e pe o med wi h S a a so - wa e, e sion 15.1 (S a aCo p, College S a ion, TX). N. Mo i e al. |3o 10 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 Resul s Nes ed Case-Con ol S udy We included 486 cases and 490 con ols om he EPIC popula- ion and 26 cases and 26 con ols om he NSHDS popula ion, wi h a median ollow-up ime o 13.9 yea s. No subs an ial di - e ences in baseline cha ac e is ics we e ound be ween cases and con ols in bo h coho s (Table 1). The baseline cha ac e is- ics o cases and con ols a e p esen ed in Supplemen a y Table 1 (a ailable online). We ound s ong co ela ions be ween log 2 - ans o med concen a ions o es one and es adiol ( ¼0.81) and and os enedione and DHEA concen a ions ( ¼0.79). Rela i ely s ong co ela ions we e obse ed be ween concen a ions o es one and and os enedione ( ¼0.5), es os- e one and and os enedione ( ¼0.58), and p oges e one and an- d os enedione ( ¼0.54) (Table 2). In he mul i a iable model, a nons a is ically signi ican ly posi i e ela ionship was ound be ween ci cula ing es one le - els and colon cance isk (odds a io [OR] pe log 2 1-uni inc emen ¼1.17 [95% CI ¼1.00 o 1.38; OR qua ile4-qua ile1[q4- q1] ¼1.33 [95% CI ¼0.89 o 1.97], P end ¼.20) (Table 3). We ound no associa ions be ween ci cula ing concen a ions o es adiol (OR q4-q1 ¼1.04 [95% CI ¼0.70 o 1.56], P end ¼.98), ee es adiol (OR q4-q1 ¼1.05 [95% CI ¼0.71 o 1.57], P end ¼.90), es os e one (OR q4-q1 ¼1.17 [95% CI ¼0.81 o 1.68], P end ¼.44), ee es os e - one (OR q4-q1 ¼1.25 [95% CI ¼0.87 o 1.79], P end ¼0.32), and os ene- dione (OR q4-q1 ¼1.12 [95% CI ¼0.77 o 1.62], P end ¼.42), DHEA (OR q4-q1 ¼0.85 [95% CI ¼0.58 o 1.23], P end ¼.78), p oges e one (OR q4-q1 ¼1.03 [95% CI ¼0.72 o 1.48], P end ¼.94), and SHBG (OR q4- q1 ¼1.00 [95% CI ¼0.69 o 1.45], P end ¼.91) and colon cance isk, wi h simila ela ionships also ound in he con inuous models (Table 3). Howe e , he posi i e ela ionship be ween ci cula ing es one le els and colon cance isk was nons a is ically signi ican a e alse-disco e y a e co ec ion. Simila associa ions we e ob- se ed when ci cula ing es one, es adiol, and es os e one mod- els we e addi ionally adjus ed o SHBG concen a ions and ice e sa (Table 3). In addi ion, we ound no e idence o an associa ion be ween he ci cula ing es adiol- o- es os e one a io and colon cance isk (OR q4-q1 ¼1.07 [95% CI ¼0.72 o 1.61], P end ¼.96). Also, a simila pa e n o esul s was ound a e he mul iple impu a ion o ho mone concen a ions wi h lowe LLOQ alues (Supplemen a y Table 2, a ailable online). We ound a simila pa - e n o associa ions when cases diagnosed wi hin he i s 2 yea s o ollow-up we e excluded om he analyses (Supplemen a y Table 3, a ailable online). In analyses by colon subsi e, he e was a nons a is ically sig- ni ican ly posi i e ela ionship be ween ci cula ing concen a- ions o es one and p oximal colon cance in he con inuous model only (OR pe log 2 1-uni inc emen ¼1.22 [95% CI ¼1.00 o 1.48]), wi h no associa ion ound o dis al colon cance (OR pe log 2 1-uni inc emen ¼1.00 [95% CI ¼0.71 o 1.41]). O he ci cu- la ing concen a ions o sex ho mones we e no associa ed wi h p oximal o dis al colon cance isk (Supplemen a y Table 4, a ailable online). Table 4 shows he esul s o subg oup analyses acco ding o BMI ca ego ies (<25 kg/m 2 and 25 kg/m 2 ). None o he associa- ions o sex ho mones wi h colon cance isk di e ed acco ding o BMI ca ego ies (all P in e ac ion .07), bu ci cula ing es one concen a ions we e posi i ely associa ed wi h colon cance isk among he o e weigh /obese g oup (BMI 25 kg/m 2 ) (OR pe log 2 inc emen ¼1.33 [95% CI ¼1.01 o 1.75) bu no he no mal- weigh g oup (OR pe log 2 inc emen ¼1.05 [95% CI ¼0.66 o 1.68], P in e ac ion ¼.07). We ound a simila pa e n o associa ions acco ding o ollow-up ime g oup (all P in e ac ion .07) (Supplemen a y Table 5, a ailable online). Me a-analysis We iden i ied 74 a icles, wi h an addi ional 6 a icles based on sc eening o i les o abs ac s. O he 80 a icles, 7 nes ed case- con ol s udies (8-12,14), including he cu en s udy, we e eligi- ble o inclusion in he me a-analysis o colo ec al, colon, and ec al cance . We pe o med 5 me a-analyses: 1) es adiol and colo ec al cance (including 6 s udies) (8-12,14), 2) es adiol and colon cance (including 2 s udies, 1 o which is he cu en s udy) (12), 3) es one and colo ec al cance (including 4 s udies), 4) es one and colon cance (including 3 s udies, 1 o which is he cu en s udy) (10,12), and 5) es one and ec al cance (in- cluding 2 s udies) (10,12)(Supplemen a y Table 6, a ailable online). In he dose- esponse me a-analysis, we ound no e idence o an associa ion be ween ci cula ing es adiol concen a ions and colo ec al cance isk (summa y RR pe 5-pg/mL inc ease in es adiol concen a ion ¼1.03 [95% CI ¼0.93 o 1.14]), wi h low he e ogenei y (I 2 ¼31.1%, P¼.20) (Figu e 1). We also ound no as- socia ion be ween ci cula ing es one concen a ions and colo- ec al cance isk (summa y RR o a 10-pg/mL inc ease in es one concen a ion ¼0.99 [95% CI ¼0.88 o 1.12]), wi h ela- i ely high he e ogenei y (I 2 ¼75.0%, P¼.007). Fo he subsi es, a nons a is ically signi ican in e se associa ion was ound be- ween ci cula ing es adiol concen a ions and colon cance isk (summa y RR o 5-pg/mL inc ease in es adiol concen- a ion ¼0.88 [95% CI ¼0.74 o 1.04]), wi h low he e ogenei y (I 2 ¼1.9%, P¼.30). No e idence o an associa ion be ween ci cu- la ing es one concen a ions and colon cance isk (summa y RR o 10-pg/mL inc ease in es one concen a ion ¼1.04 [95% CI ¼0.80 o 1.37]), wi h high he e ogenei y de ec ed (I 2 ¼81.7%, P¼.004). Li le e idence o an associa ion was ound be ween ci cula ing es one concen a ions and ec al cance isk (sum- ma y RR o 10-g/mL inc ease in es one concen a ion ¼1.05 [95% CI ¼0.82 o 1.34]), wi h low he e ogenei y (I 2 ¼0.0%, P¼.50). (Figu e 1). We we e unable o conduc a me a-analysis o he associa ion be ween ci cula ing es adiol concen a ions and ec al cance isk because only 1 s udy was a ailable (12). Discussion In his p ospec i e s udy o pos menopausal Eu opean women, we ound limi ed e idence o associa ions be ween ci cula ing concen a ions o sex ho mones and SHBG wi h colon cance isk. Associa ions we e gene ally simila when he analyses we e s a i ied acco ding o ollow-up ime and o dis al and p oximal colon cance . Simila ly, we ound no clea associa ions be ween ci cula ing es adiol and es one wi h isks o colo ec- al, colon, and ec al cance when we conduc ed dose- esponse me a-analyses encompassing all a ailable p ospec i e da a (in- cluding esul s om he cu en s udy). Consis en wi h he lowe colon cance isk obse ed o HT use s, se e al sou ces o expe imen al da a sugges ha es o- gens may con e an i umo igenic e ec s on colon umo igene- sis. I has been shown, o example, ha exp ession o es ogen ecep o -b esul s in he inhibi ion o p oli e a ion and G1 phase cell-cycle a es in colon cance cells; in xenog a mouse s ud- ies, es ogen ecep o -binhibi s cMyc exp ession and umo g ow h (35). P e iously, in he WHI-CT, we obse ed obus in- e se associa ions be ween ci cula ing es adiol and es one 4o 10 | JNCI Cance Spec um, 2021, Vol. 5, No. 6 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 wi h colo ec al and colon cance isks (12). In con as , he esul s om he cu en s udy sugges a bo de line posi i e as- socia ion be ween ci cula ing es one and cance o he colon, pa icula ly in he p oximal egion. This esul is somewha consis en wi h a s udy conduc ed wi hin he Women’s Heal h Ini ia i e Obse a ional S udy, which epo ed a posi i e associ- a ion be ween es adiol le els and colo ec al cance isk (9). O he published da a (8,10-12,14), howe e , did no obse e associa ions be ween endogenous es ogens and colon cance isk. Because o hese inconsis en indings and he ela i ely small size o each p ospec i e s udy ha has examined es o- gens and colo ec al cance in pos menopausal women, we con- duc ed a me a-analysis o combine hese da a. Impo an ly, esul s om his me a-analysis ound no e idence o an associ- a ion be ween p ediagnos ic es ogen concen a ions and colo- ec al, colon, o ec al cance isk in pos menopausal women. Table 1. Baseline cha ac e is ics o colon cance cases and con ols in EPIC and NSHDS pa icipan s a Va iable EPIC NSHDS Cases (n ¼486) Con ols (n ¼490) Cases (n ¼26) Con ols (n ¼26) Mean (SD) age a blood collec ion, y 62.0 (5.4) 61.9 (5.4) 60.4 (2.1) 60.3 (2.0) Mean (SD) body weigh , kg 68.6 (11.8) 67.5 (10.5) 75.9 (12.9) 69.5 (13.4) BMI, No. (%) 26.6 (4.6) 26.5 (4.2) 28.4 (5.5) 26.6 (4.9) Unde weigh (>18.5 kg/m 2 ) 5 (1.0) 5 (1.0) 0 (0.0) 1 (3.9) No mal (18.5-24.9 kg/m 2 ) 196 (40.3) 190 (38.8) 8 (30.8) 12 (46.2) O e weigh (25.0-29.9 kg/m 2 ) 203 (39.1) 203 (41.4) 11 (42.3) 7 (26.9) Obese (30 kg/m 2 ) 92 (19.6) 92 (18.8) 7 (26.9) 6 (23.1) Smoking s a us, No. (%) Ne e 299 (61.5) 309 (63.1) 10 (38.5) 14 (53.9) Fo me 111 (22.8) 104 (21.2) 12 (46.2) 7 (26.9) Cu en 68 (14.0) 71 (14.5) 4 (15.4) 5 (19.2) Unknown 8 (1.7) 6 (1.2) N/A N/A Physical ac i i y, No. (%) Inac i e 169 (34.8) 172 (35.1) 4 (15.4) 9 (34.6) Mode a ely inac i e 151 (31.1) 150 (30.6) 6 (23.1) 6 (23.1) Mode a ely ac i e 87 (17.9) 88 (18.0) 6 (23.1) 4 (15.4) Ac i e 65 (13.4) 68 (13.9) 8 (30.8) 5 (19.2) Missing 14 (2.9) 12 (2.5) 2 (7.7) 2 (7.7) E e used menopausal HT, No. (%) No 387 (79.6) 392 (80.0) 26 (100.0) 26 (100.0) Yes 70 (14.4) 72 (14.7) N/A N/A Unknown/missing 29 (6.0) 26 (5.3) N/A N/A Mean (SD) alcohol consump ion, g/d 5.7 (9.8) 5.9 (10.2) 3.0 (3.5) 1.5 (2.1) Se ologic a iables, median (IQR) Es one, pg/mL 18.1 (12.3-25.1) 17.7 (12.9-23.6) 25.1 (22.0-40.5) 23.9 (20.9-33.6) Es adiol, pg/mL 3.9 (2.6-6.0) 4.0 (2.6-6.0) 6.2 (4.3-11.6) 5.9 (5.1-11.5) Tes os e one, pg/mL 185.9 (129.0-257.2) 183.5 (127.9-246.2) 227.4 (166.8-298.8) 213.1 (162.4-275.8) And os enedione, ng/mL 490.1 (346.1-660.4) 466.9 (348.6-641.7) 709.0 (517.9-878.4) 566.7 (497.4-832.9) DHEA, ng/mL 1.9 (1.2-2.8) 1.8 (1.2-2.9) 2.5 (1.5-3.7) 2.4 (1.8-3.6) P oges e one, pg/mL 52.2 (37.5-76.1) 53.1 (39.6-73.5) 65.3 (51.3-100.8) 62.3 (43.5-83.0) SHBG, nmol/L 54.2 (38.4-74.6) 52.9 (40.4-70.1) 51.8 (37.5-68.8) 65.4 (52.1-86.9) F ee es adiol, pg/mL 85.8 (53.7-149.5) 91.8 (56.8-138.3) 159.0 (102.0-266.6) 125.3 (95.7-251.0) F ee es os e one, ng/mL 5.2 (3.2-7.8) 5.2 (3.3-7.4) 7.0 (4.7-9.9) 5.5 (3.4-8.9) a BMI¼body mass index; DHEA ¼dehyd oepiand os e one; EPIC ¼Eu opean P ospec i e In es iga ion in o Cance and Nu i ion; HT¼ho mone he apy; IQR ¼in e qua ile ange; N/A ¼no applicable; NSHDS ¼TheNo he nSwedenHeal handDiseaseS udy;SD¼s anda d de ia ion; SHBG ¼sex ho mone–binding p o ein. Table 2. Pea son co ela ion ma ix o ci cula ing sex ho mone concen a ions, SHBG, and BMI adjus ed o age and ba ch a Concen a ions o sex ho mone, SHBG, and BMI Es one Es adiol Tes os e one And os enedione DHEA P oges e one SHBG BMI Es one – – – – – – – – Es adiol 0.81 – – – – – – – Tes os e one 0.40 0.38 – – – – – – And os enedione 0.50 0.33 0.58 – – – – – DHEA 0.35 0.20 0.42 0.79 – – – – P oges e one 0.24 0.15 0.38 0.54 0.42 – – – SHBG –0.09 –0.19 0.14 –0.09 –0.02 0.11 – – BMI 0.28 0.39 0.08 0.07 –0.02 –0.08 –0.37 – a Se ologic a iables we e log 2 ans o med. BMI ¼body mass index; DHEA ¼dehyd oepiand os e one; SHBG ¼sex ho mone–binding globulin. N. Mo i e al. |5o 10 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 Table 3. Associa ions be ween ci cula ing concen a ions o sex ho mones and SHBG wi h colon cance in pos menopausal women (n ¼1028) Va iables Qua ile 1 Qua ile 2 Qua ile 3 Qua ile 4 P enda Con inuous model b FDR (Q alue) Es one Qua ile cu poin s <13.1 13.1-18.1 18.1-24.0 24.0 – – – No. (cases/con ols) 128/129 118/129 111/129 155/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.92 (0.64 o 1.31) 0.88 (0.61 o 1.27) 1.25 (0.87 o 1.81) .26 1.16 (0.99 o 1.35) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.94 (0.65 o 1.36) 0.91 (0.62 o 1.32) 1.32 (0.89 o 1.96) .23 1.17 (1.00 o 1.38) 0.49 Mul i a iable-adjus ed OR (95% CI) c,e 1 [Re e en ] 0.95 (0.65 o 1.37) 0.92 (0.63 o 1.35) 1.33 (0.89 o 1.97) .20 1.17 (1.00 o 1.38) – Es adiol Qua ile cu poin s <2.6 2.6-4.1 4.1-6.1 6.1 – – – No. (cases/con ols) 128/129 133/129 116/129 135/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 1.03 (0.72 o 1.47) 0.90 (0.63 o 1.28) 1.04 (0.73 o 1.50) .97 1.03 (0.92 o 1.15) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 1.03 (0.72 o 1.47) 0.89 (0.61 o 1.29) 1.04 (0.70 o 1.56) .98 1.04 (0.92 o 1.17) 0.89 Mul i a iable-adjus ed OR (95% CI) c,e 1 [Re e en ] 1.04 (0.73 o 1.48) 0.91 (0.62 o 1.32) 1.08 (0.72 o 1.61) .86 1.04 (0.92 o 1.18) – Tes os e one Qua ile cu poin s <129.3 129.3-184.5 184.5-249.9 249.9 – – – No. (cases/con ols) 126/129 122/129 117/129 147/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.98 (0.70 o 1.38) 0.93 (0.65 o 1.34) 1.18 (0.83 o 1.68) .41 1.02 (0.86 o 1.22) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.96 (0.68 o 1.36) 0.92 (0.64 o 1.33) 1.17 (0.81 o 1.68) .44 1.02 (0.86 o 1.22) 0.89 Mul i a iable-adjus ed OR (95% CI) c,e 1 [Re e en ] 0.95 (0.67 o 1.35) 0.90 (0.63 o 1.31) 1.14 (0.79 o 1.65) .51 1.01 (0.85 o 1.21) – And os enedione Qua ile cu poin s <353.8 353.8-477.7 477.7-645.5 645.5 – – – No. (cases/con ols) 136/129 104/129 123/129 149/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.77 (0.53 o 1.11) 0.91 (0.65 o 1.29) 1.11 (0.77 o 1.60) .40 1.04 (0.86 o 1.26) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.78 (0.54 o 1.14) 0.92 (0.65 o 1.31) 1.12 (0.77 o 1.62) .42 1.03 (0.85 o 1.25) 0.89 DHEA Qua ile cu poin s <1.2 1.2-1.8 1.8-2.9 2.9 – – – No. (cases/con ols) 134/129 112/129 149/129 117/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.85 (0.60 o 1.19) 1.13 (0.80 o 1.59) 0.87 (0.60 o 1.25) .84 0.98 (0.86 o 1.13) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.82 (0.58 o 1.17) 1.12 (0.79 o 1.59) 0.85 (0.58 o 1.23) .78 0.98 (0.85 o 1.12) 0.89 P oges e one Qua ile cu poin s <39.6 39.6-53.6 53.6-73.6 73.6 – – – No. (cases/con ols) 139/129 124/128 106/130 143/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.90 (0.63 o 1.27) 0.76 (0.53 o 1.08) 1.03 (0.72 o 1.47) .90 0.95 (0.80 o 1.12) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.90 (0.63 o 1.28) 0.76 (0.53 o 1.09) 1.03 (0.72 o 1.48) .94 0.95 (0.80 o 1.13) 0.89 SHBG Qua ile cu poin s <40.8 40.8-53.5 53.5-70.7 70.7 – – – No. (cases/con ols) 146/128 108/130 111/129 147/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.71 (0.50 o 1.02) 0.74 (0.52 o 1.06) 0.99 (0.71 o 1.40) .93 0.99 (0.83 o 1.19) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.72 (0.50 o 1.04) 0.75 (0.52 o 1.09) 1.00 (0.69 o 1.45) .91 0.99 (0.81 o 1.21) 0.92 Mul i a iable-adjus ed OR (95% CI) c,e 1 [Re e en ] 0.73 (0.50 o 1.05) 0.77 (0.53 o 1.12) 0.99 (0.68 o 1.44) .96 0.99 (0.81 o 1.20) – F ee es adiol Qua ile cu poin s <57.7 57.7-94.7 94.7-142.3 142.3 – – – No. (cases/con ols) 136/129 135/129 95/129 146/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.98 (0.70 o 1.37) 0.67 (0.46 o 0.98) 1.04 (0.73 o 1.46) .88 1.02 (0.92 o 1.13) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.96 (0.68 o 1.36) 0.66 (0.45 o 0.99) 1.05 (0.71 o 1.57) .90 1.03 (0.92 o 1.15) 0.89 F ee es os e one Qua ile cu poin s <3.3 3.3-5.2 5.2-7.4 7.4 – – – No. (cases/con ols) 126/129 123/129 108/129 155/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 0.99 (0.70 o 1.39) 0.87 (0.61 o 1.24) 1.25 (0.88 o 1.78) .31 1.02 (0.92 o 1.13) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 0.98 (0.69 o 1.39) 0.86 (0.60 o 1.24) 1.25 (0.87 o 1.79) .32 1.03 (0.92 o 1.15) 0.89 Es adiol- o- es os e one a io Qua ile cu poin s <0.19 0.19-0.27 0.27-0.33 0.33 – – – No. (cases/con ols) 124/129 146/129 105/129 137/129 – – – Unadjus ed OR (95% CI) c 1 [Re e en ] 1.15 (0.82 o 1.60) 0.83 (0.57 o 1.20) 1.08 (0.76 o 1.55) .97 1.21 (0.52 o 2.83) – Mul i a iable-adjus ed OR (95% CI) c,d 1 [Re e en ] 1.13 (0.81 o 1.60) 0.83 (0.56 o 1.22) 1.07 (0.72 o 1.61) .96 1.28 (0.51 o 3.24) 0.89 a S a is ical es s o end (2-sided) we e calcula ed using o dinal qua ile a iables en e ed in o he model as a single con inuous a iable. CI¼con idence in e als; DHEA ¼dehyd oepiand os e one; FDR ¼ alse-disco e y a e; OR ¼odds a io; SHBG ¼sex ho mone–binding globulin. b Se ologic a iables we e log 2 ans o med in con inuous models. c Odds a ios and 95% con idence in e als we e es ima ed by condi ional logis ic eg ession models. d Adjus ed o body mass index (unde weigh , no mal, o e weigh , o obese), smoking s a us (ne e , o me , cu en , o unknown), physical ac i i y (inac i e, mode - a ely inac i e, mode a ely ac i e, ac i e, o unknown), e e used ho mone he apy (yes, no, o unknown/missing), and alcohol consump ion (con inuous). e Addi ionally adjus ed o es one, es adiol, and es os e one o SHBG and ice e sa. 6o 10 | JNCI Cance Spec um, 2021, Vol. 5, No. 6 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 Table 4. Associa ions be ween ci cula ing concen a ions o sex ho mones and SHBG wi h colon cance in pos menopausal women (n ¼1028), by s a a o BMI Va iable Te ile 1 Te ile 2 Te ile 3 P enda Con inuous model b P alue c P in e ac iond Es one Te ile cu poin s <14.5 14.5-21.7 21.7 – – – – No. (cases/con ols) 161/171 161/173 190/172 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .07 <25 kg/m 2 1 [Re e en ] 0.83 (0.39 o1.73) 1.07 (0.36 o 3.16) .91 1.05 (0.66 o 1.68) .82 – 25 kg/m 2 1 [Re e en ] 0.92 (0.51 o 1.64) 1.47 (0.83 o 2.62) .15 1.33 (1.01 o 1.75) .04 – Es adiol Te ile cu poin s <3.0 3.0-5.2 5.2 – – – – No. (cases/con ols) 169/171 177/173 166/172 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .15 <25 kg/m 2 1 [Re e en ] 1.57 (0.71 o 3.45) 0.75 (0.27 o 2.02) .83 0.89 (0.64 o 1.25) .51 – 25 kg/m 2 1 [Re e en ] 1.16 (0.68 o 1.97) 1.26 (0.68 o 2.33) .46 1.10 (0.87 o 1.39) .41 – Tes os e one Te ile cu poin s <149.5 149.5-221.1 221.1 – – – – No. (cases/con ols) 170/171 147/172 195/173 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .51 <25 kg/m 2 1 [Re e en ] 0.71 (0.32 o 1.57) 0.97 (0.39 o 2.40) .91 0.91 (0.57 o 1.45) .69 – 25 kg/m 2 1 [Re e en ] 0.91 (0.54 o 1.52) 1.21 (0.71 o 2.06) .48 1.08 (0.81 o 1.44) .59 – And os enedione Te ile cu poin s <394.7 394.7-589.2 589.2 – – – – No. (cases/con ols) 165/171 166/171 181/173 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .12 <25 kg/m 2 1 [Re e en ] 0.56 (0.27 o 1.16) 0.77 (0.36 o 1.66) .42 0.85 (0.54 o 1.35) .50 – 25 kg/m 2 1 [Re e en ] 1.07 (0.63 o 1.84) 1.23 (0.73 o 2.07) .44 1.14 (0.84 o 1.55) .41 – DHEA Te ile cu poin s <1414.3 1414.3-2482.3 2482.3 – – – – No. (cases/con ols) 171/171 173/172 168/173 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .23 <25 kg/m 2 1 [Re e en ] 0.27 (0.11 o 0.66) 0.53 (0.23 o 1.21) .17 0.84 (0.62 o 1.16) .29 – 25 kg/m 2 1 [Re e en ] 1.15 (0.69 o 1.93) 0.96 (0.56 o 1.63) .91 1.01 (0.81 o 1.26) .91 – P oges e one Te ile cu poin s <44.1 44.1-65.4 65.4 – – – – No. (cases/con ols) 170/172 168/172 174/172 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .86 <25 kg/m 2 1 [Re e en ] 0.50 (0.24 o 1.06) 1.29 (0.56 o 2.95) .73 1.04 (0.68 o 1.59) .86 – 25 kg/m 2 1 [Re e en ] 1.27 (0.73 o 2.21) 1.27 (0.74 o 2.16) .39 1.01 (0.75 o 1.36) .95 – SHBG Te ile cu poin s <46.0 46.0-63.7 63.7 – – – – No. (cases/con ols) 195/172 131/172 186/172 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .73 <25 kg/m 2 1 [Re e en ] 0.75 (0.26 o 2.17) 0.95 (0.37 o 2.39) .91 1.31 (0.73 o 2.36) .36 – 25 kg/m 2 1 [Re e en ] 0.70 (0.41 o 1.19) 1.04 (0.59 o 1.84) .94 0.94 (0.68 o 1.30) .69 – F ee es adiol Te ile cu poin s <69.5 69.5-123.4 123.4 – – – – No. (cases/con ols) 185/171 152/172 175/173 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .07 <25 kg/m 2 1 [Re e en ] 1.21 (0.53 o 2.77) 0.51 (0.19 o 1.39) .29 0.85 (0.62 o 1.16) .30 – 25 kg/m 2 1 [Re e en ] 0.74 (0.43 o 1.28) 1.07 (0.60 o 1.91) .70 1.09 (0.88 o 1.35) .46 – F ee es os e one Te ile cu poin s <3.9 3.9-6.6 6.6 – – – – No. (cases/con ols) 157/171 161/172 194/173 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .55 <25 kg/m 2 1 [Re e en ] 0.99 (0.47 o 2.11) 1.11 (0.50 o 2.47) .81 0.85 (0.62 o 1.16) .30 – 25 kg/m 2 1 [Re e en ] 1.07 (0.63 o 1.83) 1.23 (0.72 o 2.10) .45 1.09 (0.88 o 1.35) .46 – (con inued) N. Mo i e al. |7o 10 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 In ou nes ed case-con ol s udy, we ound a posi i e associ- a ion be ween ci cula ing es one concen a ions and colon cance o o e weigh /obese women bu no o no mal-weigh women. This he e ogenei y acco ding o BMI g oup, howe e , did no each he h eshold o s a is ical signi icance, and p io s udies ha e ound limi ed e idence ha he sex ho mone–co- lo ec al cance associa ion may di e acco ding o body size (10). Gi en he mul iple s a is ical analyses conduc ed, i is pos- sible ha he posi i e associa ion we obse ed be ween es one and colon cance o o e weigh /obese women was a chance inding. Fu he s udies a e needed o examine he ole o body size on he sex ho mone–colo ec al cance associa ion. P e iously, we epo ed a posi i e associa ion be ween ci cu- la ing es os e one le els and colo ec al cance isk in a Japanese popula ion (14). I should be no ed, howe e , ha his s udy used a less sensi i e assay o measu e sex ho mone con- cen a ions, and mo e han 60% o he pa icipan s had mea- su ed es os e one concen a ions below he LLOQ (14). In he cu en analysis o pos menopausal Eu opean women, simila o s udies o UK and US women (8,36) and a ecen Mendelian andomiza ion s udy (37), we ound no associa ion be ween ci - cula ing es os e one le els and colon cance isk. We also ound no associa ion be ween ci cula ing concen a ions o an- d os enedione and DHEA wi h colon cance isk. Taken o- ge he , hese esul s p o ide li le suppo o and ogens ha ing a p ominen ole in colon cance de elopmen o pos meno- pausal women. SHBG is a hepa ically de i ed glycop o ein and p incipal anspo p o ein o es ogens and and ogens and is he e o e an impo an egula o o hei bioac i i y (38). In addi ion, Table 4. (con inued) Va iable Te ile 1 Te ile 2 Te ile 3 P enda Con inuous model b P alue c P in e ac iond Es adiol- o- es os e one a io Te ile cu poin s <0.22 0.22-0.31 0.31 – – – – No. (cases/con ols) 168/171 175/172 169/173 – – – – BMI, mul i a iable-adjus ed OR (95% CI) e .20 <25 kg/m 2 1 [Re e en ] 1.24 (0.60 o 2.54) 0.67 (0.24 o 1.87) .69 0.42 (0.05 o 3.93) .45 – 25 kg/m 2 1 [Re e en ] 1.06 (0.61 o 1.82) 1.43 (0.76 o 2.71) .26 2.10 (0.37 o 11.87) .40 a S a is ical es s o end (2-sided) we e calcula ed using o dinal e ile a iables en e ed in o he model as a single con inuous a iable. BMI ¼body mass index; CI ¼con idence in e al; DHEA ¼dehyd oepiand os e one; OR ¼odds a io; SHBG ¼sex ho mone–binding globulin. b Se ologic a iables we e log 2 ans o med in a con inuous model. c S a is ical es s (2-sided) we e calcula ed as a con inuous a iable. d He e ogenei y by BMI ca ego ies we e es ed using 2 es s. The es was 2-sided. e Odds a ios and 95% con idence in e als we e es ima ed by condi ional logis ic eg ession models. Models we e adjus ed o BMI (unde weigh , no mal, o e weigh , o obese), smoking s a us (ne e , o me , cu en , o unknown), physical ac i i y (inac i e, mode a ely inac i e, mode a ely ac i e, ac i e, o unknown), e e use ho - mone he apy (yes, no, o unknown/missing), and alcohol consump ion (con inuous). Es one, es adiol, and es os e one we e adjus ed o SHBG and ice e sa. Rela i e Risk .25 .5 .75 1 1.5 2 3 S udy Rela i e Risk (95% CI) Mo i, 2019 1.14 ( 0.99, 1.31) Falk, 2015 1.00 ( 0.85, 1.17) Mu phy, 2015 0.86 ( 0.71, 1.03) Lin, 2013 1.04 ( 0.72, 1.50) Clendenen, 2009 0.85 ( 0.47, 1.54) Gun e , 2010 1.13 ( 0.96, 1.34) O e all 1.03 ( 0.93, 1.14) A O e all (I 2= 31.1%, P=0.20) B O e all (I 2= 1.9%, P=0.30) D O e all (I 2= 81.7%, P=0.004) Rela i e Risk .25 .5 .75 1 1.5 2 3 S udy Rela i e Risk (95% CI) Falk, 2015 1.01 ( 0.97, 1.05) Mu phy, 2015 0.85 ( 0.77, 0.95) Lin, 2013 1.06 ( 0.87, 1.28) Clendenen, 2009 1.30 ( 0.92, 1.83) O e all 0.99 ( 0.88, 1.12) C O e all (I 2= 75.0%, P= 0.007) Rela i e Risk .25 .5 .75 1 1.5 2 3 S udy Rela i e Risk (95% CI) Mu phy, 2015 1.02 ( 0.79, 1.32) Clendenen, 2009 1.37 ( 0.61, 3.10) O e all 1.05 ( 0.82, 1.34) E O e all (I 2= 0.0%, P= 0.50) Mo i e al. 2019 (14) Falk e al. 2015 (11) Mu phy e al. 2015 (12) Lin e al. 2013 (8) Clendenen e al. 2009 (10) Gun e e al. 2010 (9) O e all S udy S udy O e all Rela i e Risk (95% CI) S udy O e all Mo i e al. 2021 Mu phy e al. 2015 (12) Clendenen e al. 2009 (9) Rela i e Risk (95% CI) S udy S udy Rela i e Risk (95% CI) O e all Rela i e Risk (95% CI) O e all Mo i e al. 2021 Mu phy e al. 2015 (12) Mu phy e al. 2015 (12) Clendenen e al. 2009 (10) Falk e al. 2015 (11) Mu phy e al. 2015 (12) Lin e al. 2013 (8) Clendenen e al. 2009 (10) Rela i e Risk (95% CI) Figu e 1. Dose- esponse analysis be ween ci cula ing es adiol, es one, and colo ec al cance isk. A) Es adiol and colo ec al cance , pe 5 pg/mL. B) Es adiol and co- lon cance , pe 5 pg/mL. C) Es one and colo ec al cance , pe 10 pg/mL. D) Es one and colon cance , pe 10 pg/mL. E) Es one and ec al cance , pe 10 pg/mL. The a e - age o he na u al loga i hm o he ela i e isks was es ima ed, and he ela i e isk om each s udy was weigh ed using andom e ec s weigh ing. A 2- ailed P<.05 was conside ed s a is ically signi ican . He e ogenei y be ween s udies was quan i a i ely assessed by he Q es and I 2 .The black squa es ep esen he odds a ios o he indi idual s udies and he e o ba s hei 95% con idence in e als (CIs). The a ea o he black squa es e lec s he weigh each ial con ibu es in he me a- analysis. 8o 10 | JNCI Cance Spec um, 2021, Vol. 5, No. 6 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022 SHBG is epo ed o be co ela ed wi h in lamma ion and insulin sensi i i y (39). P e iously, in he WHI-CT, we ound a mo e han 2- old highe colo ec al cance isk when he highes and lowes qua iles o SHBG we e compa ed (12). In con as , esul s om ou cu en analysis o pos menopausal Eu opean women ound no associa ion be ween SHBG concen a ions and colon cance isk, consis en wi h esul s om US, UK, and Japan-based s udies (10,14,36) as well as a ecen Mendelian andomiza ion s udy (37). O e all, hese esul s do no suppo a causal ole o SHBG in colo ec al cance de elopmen and sugges ha he p e iously epo ed associa ions we e likely bi- ased (eg, h ough con ounding) o we e he esul o chance. The cu en s udy was he la ges conduc ed o da e o ex- amine ci cula ing sex ho mone concen a ions and colon can- ce isk associa ions in pos menopausal women. A majo s eng h o he s udy was ou use o a highly sensi i e analy ical me hod o measu e sex ho mone concen a ions, wi h low in- a- and in e ba ch coe icien s o a ia ion and he b ead h o he sex ho mones we we e able o s udy. We we e also able o include impo an con ounde s, such as as ing s a us, BMI, physical ac i i y, and whe he he pa ien e e used HT. In addi- ion, we conduc ed a me a-analysis o ci cula ing es ogens and colo ec al cance and i s subsi es, including es ima es om p e- ious s udies as well as ou cu en in es iga ion. A limi a ion o he s udy was ha sex ho mone concen a ions we e mea- su ed in a single plasma sample; he e o e, hese measu e- men s may no e lec longe - e m exposu es. Such measu emen e o may no ha e been subs an ial, howe e , because a p io analysis o pos menopausal women epo ed wi hin-pe son co ela ion coe icien s anging om 0.66 o 0.92 o es one, ee es adiol, SHBG, and os enedione, es os e one, and DHEA measu emen s o e a 2- o 3-yea pe iod (40), indica - ing ha single measu es o sex ho mone concen a ions p o- ide good es ima es o longe - e m exposu es. In his p ospec i e in es iga ion o pos menopausal Eu opean women and me a-analysis o all published s udies conduc ed o da e, we ound limi ed e idence o an associa ion be ween ci cula ing concen a ions o endogenous sex ho - mones and colo ec al, colon, and ec al cance isks. Funding This wo k was suppo ed by he F ench Na ional Cance Ins i u e (INCa SHSESP17, g an No. 2017-127 o N. Mu phy). The coo dina ion o EPIC is inancially suppo ed by In e na ional Agency o Resea ch on Cance (IARC) and also by he Depa men o Epidemiology and Bios a is ics, School o Public Heal h, Impe ial College London, which has addi ional in as uc u e suppo p o ided by he NIHR Impe ial Biomedical Resea ch Cen e (BRC). The na ional coho s a e suppo ed by: Danish Cance Socie y (Denma k); Ligue Con e le Cance , Ins i u Gus a e Roussy, Mu uelle G en e ale de l’Educa ion Na ionale, Ins i u Na ional de la San  e e de la Reche che M edicale (INSERM) (F ance); Ge man Cance Aid, Ge man Cance Resea ch Cen e (DKFZ), Ge man Ins i u e o Human Nu i ion Po sdam- Rehb uecke (DI E), Fede al Minis y o Educa ion and Resea ch (BMBF) (Ge many); Associazione I aliana pe la Rice ca sul Canc o-AIRC-I aly, Compagnia di SanPaolo and Na ional Resea ch Council (I aly); Du ch Minis y o Public Heal h, Wel a e and Spo s (VWS), Ne he lands Cance Regis y (NKR), LK Resea ch Funds, Du ch P e en ion Funds, Du ch ZON (Zo g Onde zoek Nede land), Wo ld Cance Resea ch Fund (WCRF), S a is ics Ne he lands ( he Ne he lands); Heal h Resea ch Fund (FIS) - Ins i u o de Salud Ca los III (ISCIII), Regional Go e nmen s o Andaluc ıa, As u ias, Basque Coun y, Mu cia and Na a a, and he Ca alan Ins i u e o Oncology—ICO (Spain); Swedish Cance Socie y, Swedish Resea ch Council and Coun y Councils o Ska˚ne and V€ as e bo en (Sweden); Cance Resea ch UK (14136 o EPIC-No olk; C8221/A29017 o EPIC-Ox o d), Medical Resea ch Council (1000143 o EPIC-No olk; MR/ M012190/1 o EPIC-Ox o d) (Uni ed Kingdom). No es Role o he unde s: The unde s had no ole in he design o he s udy; he collec ion, analysis, and in e p e a ion o he da a; he w i ing o he manusc ip ; o he decision o submi he manusc ip o publica ion. Disclosu es: The au ho s ha e no con lic s o in e es o disclose. Au ho con ibu ions: Concep ualiza ion: SR, NMu phy, MJG; Lab analysis: PKR, AG, SR; S a is ical analysis: NMo i, DA; Supe ision: NMu phy, MJG; W i ing—o iginal d a : NMo i, SR, ND, SH, JH, AJC, NMu phy, MJG; W i ing— e iew and edi ing: all au ho s. Disclaime : Whe e au ho s a e iden i ied as pe sonnel o he In e na ional Agency o Resea ch on Cance /Wo ld Heal h O ganiza ion, he au ho s alone a e esponsible o he iews exp essed in his a icle and hey do no necessa ily ep esen he decisions, policy o iews o he In e na ional Agency o Resea ch on Cance /Wo ld Heal h O ganiza ion. Da a A ailabili y Fo in o ma ion on how o submi an applica ion o gaining ac- cess o EPIC da a and/o biospecimens, please ollow he ins uc ions a h p://epic.ia c. /access/index.php. Re e ences 1. A nold M, Sie a MS, La e sanne M, Soe joma a am I, Jemal A, B ay F. Global pa e ns and ends in colo ec al cance incidence and mo ali y. Gu . 2017; 66(4):683–691. 2. McMichael AJ, Po e JD. Rep oduc ion, endogenous and exogenous sex ho - mones, and colon cance : a e iew and hypo hesis. J Na l Cance Ins . 1980; 65(6):1201–1207. 3. Mo ch LS, Lidegaa d O, Keiding N, Lokkegaa d E, Kjae SK. The in luence o ho mone he apies on colon and ec al cance . Eu J Epidemiol. 2016;31(5): 481–489. 4. Johnson JR, Lacey JV J , Lazo ich D, e al. Menopausal ho mone he apy and isk o colo ec al cance . Cance Epidemiol Bioma ke s P e . 2009;18(1):196–203. 5. Simon MS, Chlebowski RT, Wac awski-Wende J, e al. Es ogen plus p oges in and colo ec al cance incidence and mo ali y. J Clin Oncol. 2012;30(32): 3983–3990. 6. Lin KJ, Cheung WY, Lai JY, Gio annucci EL. The e ec o es ogen s. com- bined es ogen-p oges ogen he apy on he isk o colo ec al cance . In J Cance . 2012;130(2):419–430. 7. Chlebowski RT, Wac awski-Wende J, Ri enbaugh C, e al.; Women’s Heal h Ini ia i e In es iga o s. Es ogen plus p oges in and colo ec al cance in pos menopausal women. N Engl J Med. 2004;350(10):991–1004. 8. Lin JH, Zhang SM, Rex ode KM, e al. Associa ion be ween sex ho mones and colo ec al cance isk in men and women. Clin Gas oen e ol Hepa ol. 2013; 11(4):419–424.e1. 9. Gun e MJ, Hoo e DR, Yu H, e al. Insulin, insulin-like g ow h ac o -I, endog- enous es adiol, and isk o colo ec al cance in pos menopausal women. Cance Res. 2008;68(1):329–337. N. Mo i e al. |9o 10 Downloaded om h ps://academic.oup.com/jncics/a icle/5/6/pkab084/6377340 by Uni e sidad de G anada - His o ia de las Ciencias use on 11 Feb ua y 2022