Alcohol consump ion and p os a e cance incidence
and p og ession: A Mendelian andomisa ion s udy
Clai B unne
1,2
, Neil M. Da ies
1,2
, Richa d M. Ma in
1,2,3
, Rosalind Eeles
4,5
, Doug Eas on
6
, Zso ia Ko e-Ja ai
4
,
Ali Amin Al Olama
6
, Sa a Benlloch
6
, Kenne h Mui
7
, G aham Giles
8,9
, F ed ik Wiklund
10
, Hen ik G onbe g
10
,
Ch is ophe A. Haiman
11
, Johanna Schleu ke
12,13
, Bø ge G. No des gaa d
14
, Ru h C. T a is
15
, Da id Neal
16,17
,
Jenny Dono an
1
, F eddie C. Hamdy
18
, No a Pashayan
19,20
, Kay-Tee Khaw
21
, Jane L. S an o d
22,23
, William J. Blo
24
,
S ephen Thibodeau
25
, Ch is iane Maie
26,27
, Adam S. Kibel
28,29
, Ceza y Cybulski
30
, Lisa Cannon-Alb igh
31
,
He mann B enne
32,33,34
, Jong Pa k
35
, Radka Kane a
36
, Jyo sna Ba a
37
, Manuel R. Teixei a
38
, Ha de Pandha
39
,
he PRACTICAL Conso ium,** and Luisa Zuccolo
1,2
1
School o Social and Communi y Medicine, Uni e si y o B is ol, B is ol, Uni ed Kingdom
2
MRC/Uni e si y o B is ol In eg a i e Epidemiology Uni , Uni e si y o B is ol, B is ol, Uni ed Kingdom
3
The NIHR B is ol Nu i ion Biomedical Resea ch Uni , Uni e si y Hospi als B is ol NHS Founda ion T us and he Uni e si y o B is ol, B is ol, Uni ed Kingdom
4
The Ins i u e o Cance Resea ch, London, SM2 5NG, Uni ed Kingdom
5
Royal Ma sden NHS Founda ion T us , London, SW3 6JJ, Uni ed Kingdom
Key wo ds: alcohol, p os a e cance , alcohol me abolising genes, Mendelian andomisa ion
Abb e ia ions: ADH: alcohol dehyd ogenase; ALDH: aldehyde dehyd ogenase; CI: con idence in e al; HR: haza d a io; LD: linkage
disequilib ium;; OR: odds a io; PRACTICAL: PRos a e cance AssoCia ion g oup o In es iga e Cance -associa ed AL e a ions in he
genome; SNP: single nucleo ide polymo phism
DOI: 10.1002/ijc.30436
His o y: Recei ed 14 Ap 2016; Accep ed 29 July 2016; Online 19 Sep 2016
Co espondence o: Luisa Zuccolo, MRC/Uni e si y o B is ol In eg a i e Epidemiology Uni , Uni e si y o B is ol, B is ol, Uni ed Kingdom,
Tel.: 144 [117 3310066], E-mail: [email p o ec ed]
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and
ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
G an sponso : UK MRC Special T aining Fellowship; G an numbe : G0501864/76656; G an sponso : UK MRC Popula ion Heal h
Scien is ellowship; G an numbe : G0902144; G an sponso : UK MRC; G an numbe : G0600705; G an sponso : Uni e si y o B is ol;
G an numbe : MC_UU_12013/1,9; G an sponso : The Eu opean Commission’s Se en h F amewo k P og amme; G an numbe :
HEALTH-F2-2009-223175; G an sponso : Cance Resea ch UK; G an numbe (s): C5047/A7357, C1287/A10118, C5047/A3354, C5047/
A10692, C16913/A6135, C12292/A11174, C1281/A12014, C5047/A8384, C5047/A15007, C18281/A19169; G an sponso : Na ional Ins i u es
o Heal h; G an numbe : CA128978; G an sponso : Na ional Ins i u e o Heal h (NIH) Cance Pos -Cance GWAS; G an numbe (s):
1U19 CA148537, 1U19 CA148065, 1U19 CA148112; G an sponso : Canadian Depa men o De ence; G an numbe : W81XWH-10-1-
0341; G an sponso : UK Heal h Technology Assessmen (HTA) P og amme o he NIH Resea ch; G an numbe (s): HTA 96/20/99,
ISRCTN20141297; G an sponso : Na ional Cance Resea ch Ins i u e (Depa men o Heal h, he Medical Resea ch Council and Cance
Resea ch UK); G an numbe : G0500966/75466.
The PRACTICAL Conso ium (in addi ion o hose named in he au ho lis )
In o ma ion on he conso ium can be ound a h p://p ac ical.ccge.medschl.cam.ac.uk/
Addi ional membe s om he conso ium a e Ma ga e Cook,
1
Angela Mo gan,
2
A i aya Lopha ananon,
3,4
Cy il Fishe ,
2
Daniel Leongamo n-
le ,
2
Edwa d J. Saunde s,
2
Emma J. Sawye ,
2
Ko eela Go indasami,
2
Malgo za a Tym akiewicz,
2
Michelle Guy,
2
Naomi Li ni,
2
Rosema y Wil-
kinson,
2
Sa a Jugu nau h-Li le,
2
S e e Hazel,
2
Tokhi Dadae ,
2
Melissa C. Sou hey,
5
Liesel M. Fi zge ald,
6
John Pede sen,
7
John Hoppe ,
8
Ami
Ka lsson,
9
Ca in Ca alli-Bjoe kman,
9
Jan-E ik Johansson,
9
Jan Adol son,
9
Ma kus Aly,
9,10
Michael B oms,
9
Pae S a in,
9
B ian E. Hende son,
11
F ed ick Schumache ,
11
Anssi Au inen,
12
Kimmo Taa i,
13
Liisa Maeae aenen,
14
Paula Kujala,
15
Teemu Mu ola,
16,17
Teu o L.J. Tammela,
17
Tiina Wahl o s,
18
Ma in And eas Rode ,
19
Pe e I e sen,
19
Pe e Kla sko ,
20
Sune F. Nielsen,
21,22
Tim J. Key,
23
Hans Wallinde ,
24
S en Gus a s-
son,
24
Jenny L. Dono an,
25
F eddie Hamdy,
26
Angela Cox,
27
Anne Geo ge,
28
A hene Lane,
28
Gemma Ma sden,
26
Michael Da is,
25
Paul B own,
25
Paul Pha oah,
29
Lisa B. Signo ello,
30,31
Wei Zheng,
32
Shannon K. McDonnell,
33
Daniel J. Schaid,
33
Liang Wang,
33
Lo i Tillmans,
33
Shaun
Riska,
33
Thomas Schnoelle ,
34
Ka hleen He komme ,
35
Manuel Luedeke,
34
Wal he Vogel,
36
Dominika Wokozo czyk,
37
Jan Lubiski,
37
Wojciech
Kluzniak,
37
Ka ja Bu e bach,
38
Ch is a S egmaie ,
39
Be nd Holleczek,
39
Babu Zacha iah,
40
Hui-Yi Lim,
41
Hyun Pa k,
40
James Haley,
40
Julio
Pow-Sang,
40
Ma ia Rincon,
40
Selina Radlein,
40
Thomas Selle s,
40
Cha da Sla o ,
42
Aleksand ina Vlaho a,
43
A anaska Mi ko a,
44
Da ina Kacha-
ko a,
44
Elenko Popo ,
42
S e lana Ch is o a,
43
Tihomi Diko ,
43
Vanio Mi e ,
44
Allison Ecke ,
45
Amanda Spu dle,
46
Angus Collins,
45
Glenn
Wood,
45
G eg Malone,
45
Judi h A. Clemen s,
45
K is Ke ,
45
Megan Tu ne ,
45
Pamela Saunde s,
45
Pe e Hea hco e,
45
S ilakshmi S ini asan,
45
Lei e Moya,
45
T ina Yeadon,
45
Aus alian P os a e Cance BioResou ce,
45
Joana San os,
47
Ca men Je 
onimo,
47
Paula Paulo,
47
Ped o Pin o,
47
Rui
Hen ique,
47
Sofia Maia,
47
Agnieszka Michael,
48
And zej Kie zek,
48
Huihai Wu
48
andRobe J.MacInnis
6,8
Cance Epidemiology
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
In e na ional Jou nal o Cance
IJC
6
S angeways Labo a o y, Depa men o Public Heal h and P ima y Ca e, Cen e o Cance Gene ic Epidemiology, Uni e si y o Camb idge, Wo s
Causeway, Camb idge, Uni ed Kingdom
7
Ins i u e o Popula ion Heal h, Uni e si y o Manches e , Manches e , Uni ed Kingdom
8
The Cance Council Vic o ia, Cance Epidemiology Cen e, 1 Ra hdowne S ee , Ca l on, Vic, Aus alia
9
Cen e o Molecula , En i onmen al, Gene ic and Analy ic Epidemiology, The Uni e si y o Melbou ne, Vic, Aus alia
10
Depa men o Medical Epidemiology and Bios a is ics, Ka olinska Ins i u e, S ockholm, Sweden
11
Depa men o P e en i e Medicine, Keck School o Medicine, Uni e si y o Sou he n Cali o nia/No is Comp ehensi e Cance Cen e , Los Angeles, CA
12
Depa men o Medical Biochemis y and Gene ics, Uni e si y o Tu ku, Tu ku, Finland
13
Ins i u e o Biomedical Technology/BioMediTech, Uni e si y o Tampe e and FimLab Labo a o ies, Tampe e, Finland
14
Depa men o Clinical Biochemis y, He le Hospi al, Copenhagen Uni e si y Hospi al, He le Ring ej 75, DK-2730, He le , Denma k
15
Cance Epidemiology Uni , Nu ield Depa men o Clinical Medicine, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom
16
Su gical Oncology (U o-Oncology: S4), Uni e si y o Camb idge, Box 279, Addenb ooke’s Hospi al, Hills Road, Camb idge, Uni ed Kingdom
17
Cance Resea ch UK Camb idge Resea ch Ins i u e, Li Ka Shing Cen e, Camb idge, Uni ed Kingdom
18
Nu ield Depa men o Su ge y, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom
19
S angeways Labo a o y, Depa men o Oncology, Cen e o Cance Gene ic Epidemiology, Uni e si y o Camb idge, Wo s Causeway, Camb idge, Uni ed Kingdom
20
Depa men o Applied Heal h Resea ch, Uni e si y College London, 1-19 To ing on Place, London, WC1E 7HB, Uni ed Kingdom
21
Camb idge Ins i u e o Public Heal h, Uni e si y o Camb idge, Fo ie Si e, Robinson Way, Camb idge, CB2 0SR, Uni ed Kingdom
1
Cen e o Cance Gene ic Epidemiology, Depa men o Public Heal h and P ima y Ca e, Uni e si y o Camb idge, S angeways Labo a o y,
Wo s Causeway, Camb idge CB1 8RN, Uni ed Kingdom,
2
The Ins i u e o Cance Resea ch, Su on, Uni ed Kingdom,
3
Ins i u e o Popula ion
Heal h, Uni e si y o Manches e , Manches e , Uni ed Kingdom,
4
Wa wick Medical School, Uni e si y o Wa wick, Co en y, Uni ed Kingdom,
5
Gene ic Epidemiology Labo a o y, Depa men o Pa hology, The Uni e si y o Melbou ne, G a an S ee , Pa k ille, Vic, Aus alia,
6
Cance Epide-
miology Cen e, The Cance Council Vic o ia, 615 S Kilda Road, Melbou ne, Vic, Aus alia,
7
Tissupa h P y L d., Melbou ne, Vic Aus alia,
8
Cen e
o Epidemiology and Bios a is ics, Melbou ne School o Popula ion and Global Heal h, The Uni e si y o Melbou ne, Vic, Aus alia,
9
Depa men
o Medical Epidemiology and Bios a is ics, Ka olinska Ins i u e, S ockholm, Sweden,
10
Depa men o Clinical Sciences a Dande yds Hospi al,
S ockholm, Sweden,
11
Depa men o P e en i e Medicine, Keck School o Medicine, Uni e si y o Sou he n Cali o nia/No is Comp ehensi e Can-
ce Cen e , Los Angeles, CA,
12
Depa men o Epidemiology, School o Heal h Sciences, Uni e si y o Tampe e, Tampe e, Finland,
13
Depa men o
U ology, Helsinki Uni e si y Cen al Hospi al and Uni e si y o Helsinki, Helsinki, Finland,
14
Finnish Cance Regis y, Helsinki, Finland,
15
Fimlab
Labo a o ies, Tampe e Uni e si y Hospi al, Tampe e, Finland,
16
School o Medicine, Uni e si y o Tampe e, Tampe e, Finland,
17
Depa men o
U ology, Tampe e Uni e si y Hospi al and Medical School, Uni e si y o Tampe e, Finland,
18
BioMediTech, Uni e si y o Tampe e and FimLab
Labo a o ies, Tampe e, Finland,
19
Copenhagen P os a e Cance Cen e , Depa men o U ology, Rigshospi ale , Copenhagen Uni e si y Hospi al,
Tagens ej 20, 7521, DK-2200 Copenhagen, Denma k,
20
Depa men o U ology, He le Hospi al, Copenhagen Uni e si y Hospi al, He le Ring ej
75, DK-230 He le , Denma k,
21
Depa men o Clinical Biochemis y, He le Hospi al, Copenhagen Uni e si y Hospi al, He le Ring ej 75, DK-
230 He le , Denma k,
22
Facul y o Heal h and Medical Sciences, Uni e si y o Copenhagen,
23
Cance Epidemiology Uni , Nu field Depa men o
Clinical Medicine, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom,
24
Depa men o Epidemiology and Bios a is ics, School o Public Heal h, Impe-
ial College, London, Uni ed Kingdom,
25
School o Social and Communi y Medicine, Uni e si y o B is ol, Canynge Hall, 39 Wha ley Road, B is ol,
BS8 2PS, Uni ed Kingdom,
26
Nu field Depa men o Su gical Sciences, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom, Facul y o Medical Science,
Uni e si y o Ox o d, John Radcli e Hospi al, Ox o d, Uni ed Kingdom,
27
CR- Uni ed Kingdom /YCR She field Cance Resea ch Cen e, Uni e si y
o She field, She field, Uni ed Kingdom,
28
Uni e si y o Camb idge, Depa men o Oncology, Box 279, Addenb ooke’s Hospi al, Hills Road Cam-
b idge CB2 0QQ, Uni ed Kingdom,
29
Cen e o Cance Gene ic Epidemiology, Depa men o Oncology, Uni e si y o Camb idge, S angeways
Labo a o y, Wo s Causeway, Camb idge, Uni ed Kingdom,
30
In e na ional Epidemiology Ins i u e, 1555 Resea ch Bl d., Sui e 550, Rock ille, MD,
31
Depa men o Epidemiology, Ha a d School o Public Heal h, 677 Hun ing on A enue, Bos on, MA,
32
Di ision o Epidemiology, Depa men o
Medicine, Vande bil Uni e si y Medical Cen e , 2525 Wes End A enue, Sui e 800, Nash ille, TN,
33
Mayo Clinic, Roches e , MN,
34
Depa men o
U ology, Uni e si y Hospi al Ulm, Ge many,
35
Depa men o U ology, Klinikum ech s de Isa de Technischen Uni e si ae Muenchen, Munich,
Ge many,
36
Ins i u e o Human Gene ics, Uni e si y Hospi al Ulm, Ge many,
37
In e na ional He edi a y Cance Cen e , Depa men o Gene ics
and Pa hology, Pome anian Medical Uni e si y, Szczecin, Poland,
38
Di ision o Clinical Epidemiology and Aging Resea ch, Ge man Cance
Resea ch Cen e (DKFZ), 69120 Heidelbe g, Ge many,
39
Saa land Cance Regis y, 66119 Saa b uecken, Ge many,
40
Depa men o Cance Epide-
miology, Mo fi Cance Cen e , 12902 Magnolia D i e, Tampa, FL,
41
Bios a is ics P og am, Mo fi Cance Cen e , 12902 Magnolia D i e, Tampa,
FL,
42
Depa men o U ology and Alexand o ska Uni e si y Hospi al, Medical Uni e si y, Sofia, Bulga ia,
43
Depa men o Gene al and Clinical
Pa hology, Medical Uni e si y, Sofia, Bulga ia,
44
Depa men o Medical Chemis y and Biochemis y, Molecula Medicine Cen e , Medical Uni e -
si y, Sofia, 2 Zd a e S ., 1431 Sofia, Bulga ia,
45
Aus alian P os a e Cance Resea ch Cen e-Qld, Ins i u e o Heal h and Biomedical Inno a ion and
School o Biomedical Science, Queensland Uni e si y o Technology, BNE, Aus alia,
46
Molecula Cance Epidemiology Labo a o y, Queensland
Ins i u e o Medical Resea ch, BNE, Aus alia,
47
Depa men o Gene ics, Po uguese Oncology Ins i u e, Po o, Po ugal,
48
The Uni e si y o Su ey,
Guild o d, Su ey GU2 7XH, Uni ed Kingdom
Con lic o in e es : Rosalind Eeles epo s an hono a ia o a end he Geni ou ina y Cance s Symposium a he Ame ican Socie y o Clinical
Oncology Mee ing Janua y 2016.
Cance Epidemiology
76 Alcohol me abolism and p os a e cance
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
22
Di ision o Public Heal h Sciences, F ed Hu chinson Cance Resea ch Cen e , Sea le, WA
23
Depa men o Epidemiology, School o Public Heal h, Uni e si y o Washing on, Sea le, WA
24
In e na ional Epidemiology Ins i u e, 1455 Resea ch Bl d, Sui e 550, Rock ille, MD
25
Mayo Clinic, Roches e , MN
26
Depa men o U ology, Uni e si y Hospi al Ulm, Ge many
27
Ins i u e o Human Gene ics Uni e si y Hospi al Ulm, Ge many
28
B igham and Women’s Hospi al/Dana-Fa be Cance Ins i u e, 45 F ancis S ee -ASB II-3, Bos on, MA
29
Washing on Uni e si y, S Louis, MO
30
Depa men o Gene ics and Pa hology, In e na ional He edi a y Cance Cen e , Pome anian Medical Uni e si y, Szczecin, Poland
31
Di ision o Gene ic Epidemiology, Depa men o Medicine, Uni e si y o U ah School o Medicine, Sal Lake Ci y, UT
32
Di ision o Clinical Epidemiology and Aging Resea ch, Ge man Cance Resea ch Cen e (DKFZ), Heidelbe g, Ge many
33
Di ision o P e en i e Oncology, Ge man Cance Resea ch Cen e (DKFZ), Heidelbe g, Ge many
34
Ge man Cance Conso ium (DKTK), Ge man Cance Resea ch Cen e (DKFZ), Heidelbe g, Ge many
35
Di ision o Cance P e en ion and Con ol, H. Lee Mo i Cance Cen e , 12902 Magnolia D , Tampa, FL
36
Depa men o Medical Chemis y and Biochemis y, Molecula Medicine Cen e , Medical Uni e si y So ia, 2 Zd a e S , So ia, 1431, Bulga ia
37
Aus alian P os a e Cance Resea ch Cen e-Qld, Ins i u e o Heal h and Biomedical Inno a ion and Schools o Li e Science and Public Heal h,
Queensland Uni e si y o Technology, BNE, Aus alia
38
Depa men o Gene ics, Po uguese Oncology Ins i u e, Po o, Po ugal and Biomedical Sciences Ins i u e (ICBAS), Po o Uni e si y, Po o, Po ugal
39
The Uni e si y o Su ey, Guild o d, Su ey, GU2 7XH, Uni ed Kingdom
P os a e cance is he mos common cance in men in de eloped coun ies, and is a a ge o isk educ ion s a egies. The
e ec s o alcohol consump ion on p os a e cance incidence and su i al emain unclea , po en ially due o me hodological
limi a ions o obse a ional s udies. In his s udy, we in es iga ed he associa ions o gene ic a ian s in alcohol-me abolising
genes wi h p os a e cance incidence and su i al. We analysed da a om 23,868 men wi h p os a e cance and 23,051 con-
ols om 25 s udies wi hin he in e na ional PRACTICAL Conso ium. S udy-speci ic associa ions o 68 single nucleo ide poly-
mo phisms (SNPs) in 8 alcohol-me abolising genes (Alcohol Dehyd ogenases (ADHs) and Aldehyde Dehyd ogenases (ALDHs))
wi h p os a e cance diagnosis and p os a e cance -speci ic mo ali y, by g ade, we e assessed using logis ic and Cox eg es-
sion models, espec i ely. The da a ac oss he 25 s udies we e me a-analysed using ixed-e ec and andom-e ec s models.
We ound li le e idence ha a ian s in alcohol me abolising genes we e associa ed wi h p os a e cance diagnosis. Fou a -
ian s in wo genes exceeded he mul iple es ing h eshold o associa ions wi h p os a e cance mo ali y in ixed-e ec
me a-analyses. SNPs wi hin ALDH1A2 associa ed wi h p os a e cance mo ali y we e s1441817 ( ixed e ec s haza d a io,
HR
ixed
50.78; 95% con idence in e al (95%CI):0.66,0.91; p alues 50.002); s12910509, HR
ixed
50.76; 95%CI:0.64,0.91; p
alues 50.003); and s8041922 (HR
ixed
50.76; 95%CI:0.64,0.91; p alues 50.002). These SNPs we e in linkage disequilib i-
um wi h each o he . In ALDH1B1, s10973794 (HR
ixed
51.43; 95%CI:1.14,1.79; p alues 50.002) was associa ed wi h p os-
a e cance mo ali y in men wi h low-g ade p os a e cance . These esul s sugges ha alcohol consump ion is unlikely o
a ec p os a e cance incidence, bu i may in luence disease p og ession.
In oduc ion
P os a e cance is he mos common cance in men in de el-
oped coun ies, wi h 758,700 new cases diagnosed and
142,000 dea hs in 2012.
1
Wi h inc easing up ake o p os a e-
specific an igen (PSA) es ing and he ageing popula ion,
p os a e cance incidence is inc easing.
2
The ac o s influenc-
ing p os a e cance incidence and su i al a e diagnosis a e
poo ly unde s ood, he e o e mo e e idence is needed.
3
Alcohol is a ca cinogen associa ed wi h o opha yngeal, li -
e , b eas , colo ec al and oesophageal cance s.
4
Func ional
a ia ion in he genes in ol ed in alcohol me abolism esul
in al e ed exposu e o he ca cinogenic me aboli es o e ha-
nol, sugges ing a mechanism o gene ic sensi i i y o alcohol
o influence he pa hogenesis o cance s.
5
Fo example, popu-
la ions wi h an inc eased p e alence o common gene ic a i-
a ion in he alcohol dehyd ogenase gene, ha esul s in
Wha ’s new?
Alcohol may spu p os a e cance p og ession, hough i does no appea o a ec incidence, acco ding o new analysis. Va ia-
ion in genes in ol ed in alcohol me abolism a ec how much he body is exposed o ca cinogenic me aboli es. These au ho s
examined 68 gene ic a ian s in alcohol dehyd ogenase (ADH) and aldehyde dehyd ogenase (ALDH) genes, seeking a link wi h
p os a e cance isk. While hey ound no e idence ha hese a ian s al e p os a e cance incidence, hey did show ha
SNPs in he ALDH1A2 gene a ec p os a e cance mo ali y. F om a public heal h s andpoin , hese esul s sugges educing
alcohol consump ion could slow p os a e cance disease p og ession.
Cance Epidemiology
B unne e al. 77
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
educed enzyme ac i i y, ha e an inc eased isk o oesopha-
geal cance compa ed wi h popula ions wi h he ully ac i e
enzyme.
4
A p esen he ole o alcohol use on p os a e cance
emains unce ain. The Wo ld Cance Resea ch Fund’s ex en-
si e epo based on sys ema ic e iews desc ibed he e idence
as limi ed and inconclusi e,
6
and he In e na ional Associa ion
o Resea ch on Cance did no lis his cance si e amongs
o he s mo e appa en ly caused by alcohol in hei Monog aph
on alcohol’s ca cinogenici y.
4
The e ha e been conflic ing
epo s o possible associa ions o alcohol wi h a ious s ages
o his ological g ades o p os a e cance ,
7–12
and me a-analyses
ha e highligh ed he inconsis encies, emphasising he need o
u he esea ch in his a ea.
13,14
The majo i y o e idence abou he e ec s o alcohol on
p os a e cance is om obse a ional s udies. One po en ial
limi a ion o adi ional obse a ional esea ch is ha he
findings can po en ially be explained by common causes o
bo h exposu e and ou come (con ounding ac o s). O he
po en ial sou ces o bias a e e e se causa ion and ecall bias,
whe e ha ing p os a e cance a ec s d inking beha iou o
i s epo ing, a he han alcohol consump ion inc easing he
isk o p os a e cance . A p ospec i e s udy design could mi -
iga e bo h o hese p oblems, bu could s ill be a ec ed by
bias in he o m o he “sick qui e ” e ec , whe e o me
hea y d inke s educe hei alcohol in ake in middle-age
because o como bidi ies ha may be alcohol- ela ed.
Mendelian andomisa ion is an app oach ha uses gene ic
a ian s obus ly associa ed wi h exposu es o in e es , o
hei me abolic e ec s, as ins umen al a iables o es he
un-con ounded and unbiased causal e ec s o hose expo-
su es and hei me abolic e ec s wi h cance .
15
Mendelian
andomisa ion analyses ely on wo app oxima e laws o
Mendelian gene ics,
15,16
ha a meiosis alleles seg ega e wi h-
ou any influence o en i onmen al ac o s and ha he
inhe i ance o one ai is independen o he inhe i ance o
o he s. This allows gene ic a ia ion o be used in epidemio-
logical s udies as an un-con ounded p oxy o an en i on-
men al exposu e,
15–17
in his case alcohol consump ion, o
es ima e he influence o cumula i e li e- ime isk o expo-
su e, o educe ecall bias and he “sick-qui e ” e ec and o
nega e e e se causa ion. All hese ea u es a e limi a ions o
p e ious con en ional obse a ional s udies.
7,8,13
Mendelian
andomisa ion has al eady been used success ully in bo h ca -
dio ascula
18,19
and cance epidemiology
20–24
o cla i y he
causal e ec s o alcohol on disease.
In his s udy, we unde ook Mendelian andomisa ion
analyses in which we used a ian s in alcohol me abolising
genes influencing me abolism and in ake, o es he causal
e ec o alcohol exposu e on p os a e cance isk and p o-
g ession. The mo i a ion is ha i alcohol in ake causally
inc eases p os a e cance isk o p og ession, hen gene ic
a ian s associa ed wi h me abolic e ec s o alcohol o
inc eased in ake will be di e en ially ep esen ed in cases and
con ols. We s a ified he analysis by his ological p os a e
cance g ade, based on Gleason sco e, as low- and high-g ade
p os a e cance s ha e di e ing na u al his o ies which could
be influenced by di e en isk ac o s.
Ma e ial and Me hods
S udy popula ions
We used pheno ypic and geno ypic da a om 46,919 men
(23,868 cases) in he in e na ional P os a e cance associa ion
g oup o in es iga e cance -associa ed al e a ions in he
genome (PRACTICAL) conso ium. Da a we e p o ided by
25 s udies wi hin he conso ium, based in USA, Aus alia
and Eu opean coun ies. This s udy popula ion was limi ed
o hose o Eu opean e hnici y. The s udies used a numbe o
me hods o ec ui men , including sc een and clinically
de ec ed cases and pa icipan s selec ed due o a amily his o-
y o p os a e cance . The backg ound cha ac e is ics o he
pa icipan s o each s udy a e shown in Table 1. Gleason
sco es we e used o ca ego ise cance s as low g ade (Gleason
sco e 6) o high g ade (Gleason sco e 7). Fu he de ails
a e a ailable om he conso ium websi e (p ac ical.ccge.-
medschl.cam.ac.uk). All s udies adhe ed bo h o na ional e hi-
cal guidelines and o he p inciples o he Decla a ion o
Helsinki.
Geno yping da a
The pa icipan s we e geno yped using a cus om Illumina
Infinium geno yping a ay (iCOGS), which was specifically
designed o he Collabo a i e Oncological Gene-
en i onmen al S udy (COGS) and eco ded 211,155 SNPs
(de ails a ailable om: h p://ec.eu opa.eu/ esea ch/heal h/
medical- esea ch/cance / p7-p ojec s/cogs_en.h ml).
25,26
The
iCOGS a ay was designed o in es iga e SNPs in egions
hough o be associa ed wi h b eas , o a ian and p os a e
cance ; 68,638 o he SNPs on he a ay we e chosen because
o hei po en ial ole in p os a e cance ae iology. The o he
125,877 SNPs we e selec ed on he basis o po en ial impo -
ance o o he cance s and common SNPs known o be
associa ed wi h any o he ai s. Indi iduals wi h ewe han
95% o geno ypes called, o high o low he e ozygosi y (p<1
310
25
) we e d opped om ou analysis. In o al 201,598
SNPs passed quali y con ol. The geno ypic da a we e used o
impu e SNPs which we e no di ec ly geno yped, bu we e
in linkage disequilib ium (LD) wi h geno yped SNPs. We
used he HapMap 2 CEU e e ence panel and IMPUTE2
so wa e.
27
In his s udy, we sea ched he iCOGS a ay da abase o all
alcohol-me abolising gene ic a ian s (wi hin Alcohol Dehy-
d ogenases (ADHs) o Aldehyde Dehyd ogenases (ALDHs)
genes), and iden ified 68 common a ian s in 5 dis inc geno-
mic egions: he ADH clus e on ch omosome 4, comp ising
ADH1A, ADH1B, ADH1C and ADH7; ALDH1A1 and ALDH
1B1 on ch omosome 9; ALDH1A2 and ALDH1A3 on ch omo-
some 15. O hese SNPs, 67 we e di ec ly geno yped, and
one was impu ed. An o e iew o he genes’ ole in alcohol
me abolism and beha iou is shown in Supplemen a y ma e ial
Cance Epidemiology
78 Alcohol me abolism and p os a e cance
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
Table S1. The cha ac e is ics o he SNPs included in his s udy
a e shown in Supplemen a y ma e ial Table S2.
S a is ical analysis
We con e ed he geno ypic da a o each SNP in o a coun o
he numbe o mino alleles a each locus. We used logis ic
eg ession o es ima e he associa ions o he SNPs wi h p os a e
cance isk (cases s. con ols), o all cases and s a ified by
high- s. low-g ade disease. In a case-only analysis, we used Cox
p opo ional haza ds eg ession o es ima e associa ions o each
SNP wi h p os a e cance -specific mo ali y, s a ified by g ade
o p os a e cance . All eg ession analyses we e adjus ed o he
fi s 8 p incipal componen s o popula ion s a ifica ion, since
hese genomic egions show ma ked a ia ion ac oss di e en
popula ions and so do p os a e cance incidence and su i al.
When checking he p opo ional haza ds assump ion, we ound
li le e idence o iola ion. These eg essions we e pe o med o
each s udy and hen me a-analysed using bo h fixed- and
andom-e ec s models. S udies we e excluded om he su i al
me a-analysis i he e we e <5 dea hs du ing he ollow up pe i-
od o <90% comple ion o ollow up da a. We in es iga ed
be ween-s udy he e ogenei y using he S a a me an command o
es ima e he I
2
s a is ic assuming a fixed-e ec model; we also
epo andom-e ec models o comple eness as such models
may be ele an whe e I
2
alues a e high (e.g., >75%).
28
Using
me a- eg ession, we in es iga ed whe he he e ogenei y could be
explained by he ollowing a p io i defined s udy-specific cha ac-
e is ics: mean age a diagnosis, mean PSA a diagnosis, coun y
o s udy (USA s. elsewhe e), and he pe cen age o pa icipan s
wi h a amily his o y o p os a e cance . Manha an plo s o he
associa ions in he fi e ch omosomal egions we e cons uc ed
o iden i y SNPs exceeding he Nyhol co ec ed p alues h esh-
old o associa ion —a mul iple es ing co ec ion which
accoun s o LD be ween he SNPs.
29
Sensi i i y analyses we e conduc ed by eclassi ying low- and
high-g ade disease as <8 and 8–10 Gleason g ade, espec i ely.
The powe o ou s udy was also assessed using e e se
powe calcula ions o demons a e he e ec size we would
Table 1. Backg ound in o ma ion on pa icipan s con ibu ing o he PRACTICAL Conso ium by s udy
Age a
diagnosis PSA le el a diagnosis (ng/ml) Family his o y Gleason sco e
S udy Coun y Con ols Cases Mean SD Median Lowe qua ile Uppe qua ile o disease 8–10
CAPS Sweden 664 1,153 66.10 7.75 13.0 7.0 30.0 17.35% 15.26%
CPCS1 Denma k 2,771 848 69.51 7.91 15.0 8.0 37.5 8.21% 26.65%
CPCS2 Denma k 1,009 265 64.88 6.82 9.0 6.0 14.5 14.72% 9.06%
EPIC Eu ope 1,079 722 64.87 5.62 8.6 6.0 15.9 – 2.22%
EPIC-No olk UK 917 484 72.08 7.56 19.8 19.8 19.8 2.48% 1.86%
ESTHER Ge many 318 313 65.52 5.09 6.9 5.0 14.0 10.54% 8.63%
FHCRC USA 730 761 59.73 7.18 6.4 4.7 9.8 21.68% 10.38%
IPO-Po o Po ugal 66 183 59.33 5.23 7.4 5.5 10.1 20.00% 15.85%
MAYO USA 488 767 65.24 6.42 7.8 4.9 14.7 29.07% 28.42%
MCCS Aus alia 1,170 1,698 58.45 8.46 5.4 0.0 11.4 23.45% 10.31%
MEC USA 829 819 69.53 7.62 – – – 13.03% 34.55%
MOFFITT USA 100 414 64.97 8.27 5.6 4.3 7.4 22.76% 11.11%
PCMUS Bulga ia 140 151 69.27 8.71 15.8 7.4 34.0 5.30% 29.80%
PPF-UNIS UK 188 245 68.86 7.57 8.6 6.3 14.0 25.22% 9.39%
Poland Poland 359 438 67.66 7.84 11.0 6.9 26.0 10.57% 11.42%
P oMPT UK 0 166 66.33 8.64 8.8 5.7 15.3 34.62% 16.87%
P o ecT UK 1,474 1,542 62.76 5.11 5.1 3.8 8.2 7.91% 5.64%
QLD Aus alia 87 186 61.32 6.91 5.2 2.2 7.5 36.18% 3.76%
SEARCH UK 1,244 1,371 63.08 4.76 8.8 5.6 15.0 16.24% 10.14%
STHM1 Sweden 2,224 2,006 66.17 6.99 – – – 20.18% 7.93%
TAMPERE Finland 2,413 2,754 68.18 7.96 8.6 5.6 16.3 – 13.76%
UKGPCS UK 4,182 4,549 63.76 7.97 9.8 5.6 24.7 23.42% 14.13%
ULM Ge many 354 603 63.78 6.66 9.0 6.0 15.1 44.94% 12.11%
UTAH USA 245 440 62.57 8.85 – – – 51.36% 15.45%
WUGS USA 0 990 60.80 7.03 5.0 4.0 7.0 42.43% 7.88%
Abb e ia ions: SD, s anda d de ia ion; PSA, p os a e-speci ic an igen.
Cance Epidemiology
B unne e al. 79
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC

expec o de ec gi en ou sample size and a50.05 wi h
SNPs o a ange o mino allele equencies.
30
The analysis
was ca ied ou using S a a .13.1. The s a is ical code used
o p oduce hese esul s can be accessed he e (h ps://gi hub.
com/nmda ies/p ac ical-alcohol/).
Resul s
The backg ound cha ac e is ics o he pa icipan s a e sum-
ma ised in Table 1, by s udy. Va ia ion be ween s udies
eflec s hei indi idual ec ui men me hods (e.g., some
s udies selec ed o hose men wi h a posi i e amily his o y
o p os a e cance ).
The Manha an plo s wi h esul s om bo h fixed and
andom e ec s me a-analyses es ing associa ions be ween
SNPs in he fi e genomic egions and p os a e cance isk a e
p esen ed in Figu e 1 (de ailed esul s a ailable in Supple-
men a y ma e ial Tables S4–S9, and sensi i i y analyses wi h
al e na i e defini ions o low- and high-g ade p esen ed in
Supplemen a y ma e ial Tables S16–S19). The figu e and
Supplemen a y ma e ial ables show ha no SNP exceeded
Figu e 1. Manha an plo s o associa ion o SNPs, in 5 egions in ol ed in alcohol me abolism, wi h P os a e Cance Diagnosis by P os a e
Cance g ade.
Figu e 2. Manha an plo s o associa ion o SNPs, in 5 egions in ol ed in alcohol me abolism, wi h p os a e cance -speci ic su i al by
p os a e Cance g ade.
Cance Epidemiology
80 Alcohol me abolism and p os a e cance
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
he Nyhol co ec ed p alues h eshold o associa ion wi h
p os a e cance isk.
In case-only analyses, ou SNPs exceeded he Nyhol
co ec ed p alues h eshold o associa ion wi h p os a e
cance -specific mo ali y in he fixed-e ec me a-analysis
(summa y Manha an plo s p esen ed in Figu e 2, and indi-
idual SNP esul s p esen ed in Supplemen a y ma e ial
Tables S10–S15, wi h esul s o sensi i i y analyses wi h
al e na i e defini ions o low- and high-g ade p esen ed in
Supplemen a y ma e ial Tables S20–S23). Th ee SNPs wi h-
in ALDH1A2 we e associa ed wi h p os a e cance mo ali-
y ollowing diagnosis wi h any p os a e cance : s1441817
(fixed e ec s haza d a io, HR
fixed
50.78; 95% confidence
in e al (95%CI):0.66,0.91, p alues 50.002, I
2
519.4);
Figu e 3. Me a-analysis o p os a e cance -speci ic su i al ollowing a diagnosis o any p os a e cance , in associa ion wi h s1441817 in
ALDH1A2. [Colo igu e can be iewed a wileyonlinelib a y.com]
Figu e 4. Me a-analysis o p os a e cance -speci ic su i al ollowing a diagnosis o low g ade p os a e cance , in associa ion wi h
s10973794 in ALDH1B1. [Colo igu e can be iewed a wileyonlinelib a y.com]
Cance Epidemiology
B unne e al. 81
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
s12910509, HR
fixed
50.76; 95%CI:0.64,0.91, p al-
ues 50.003, I
2
523.0); and s8041922 (HR
fixed
50.76;
95%CI:0.64,0.91, p alues 50.002, I
2
525.5). To iden i y
he op independen signal amongs hese h ee, we con-
duc ed join ly adjus ed analyses. Le els o pai wise LD we e
oo high o a emp s udy-specific analyses (pai wise LD
s1441817 and s12910509
2
50.89, s1441817 and
s8041922
2
50.88, s12910509 and s8041922
2
50.99).
31
Pooled analyses we e conduc ed o es ima e he join e ec s
o s1441817 and s12910509/ s8041922 (
2
50.99) on
p os a e cance su i al, wi h a andom e ec co ec ion
o s anda d e o s. These showed an independen e ec o
s1441817, simila in size o ha o uni a ia e analyses, bu
no independen e ec o s12910509/ s8041922 once
adjus ing o s1441817. Gi en he high LD be ween he
h ee SNPs, hey should be aken as ep esen ing one
unde lying gene ic signal. Figu e 3 p esen s he o es plo
o indi idual s udies con ibu ing o he me a-analysis o
p os a e cance -specific su i al in associa ion wi h he op
independen signal in ALDH1A2. A ou h SNP,
s10973794 in ALDH1B1 (also close o IGFBPL1), was
associa ed wi h p os a e cance mo ali y in men wi h a diag-
nosis o low-g ade p os a e cance (HR
fixed
51.43;
95%CI:1.14,1.79, p alues 50.002, I
2
523.4, Fig. 4). This esul
was obus o changing he low-g ade defini ion o
<8 (HR
fixed
51.23; 95% CI:1.06,1.41, p alues 50.002, I
2
50,
Supplemen a y ma e ial Table S22).
In gene al, andom e ec s me a-analyses yielded weake
e idence o associa ion han fixed-e ec models, as one
would expec due o a iabili y ac oss s udies p oducing la g-
e confidence in e als o he o me . Howe e , le els o he -
e ogenei y as quan ified by he I
2
s a is ics we e low, wi h
poin es ima es ema kably simila ac oss he wo ypes o
me a-analysis. Me a- eg ession analyses ound limi ed e i-
dence ha he s udy-le el cha ac e is ics examined had a
s ong influence on he pooled esul s (Table 2).
Discussion
Using da a om he PRACTICAL Conso ium, we pooled
da a om 25 s udies including a o al o 23,868 p os a e can-
ce cases and 23,091 con ols, o in es iga e he associa ion o
68 SNPs wi hin genes hough o be in ol ed wi h alcohol
me abolism wi h p os a e cance isk and p os a e cance -
specific mo ali y (amongs men diagnosed wi h p os a e can-
ce ), o e all and by Gleason g ade. A e co ec ing o mul i-
ple es ing in he fixed-e ec me a-analysis, no SNPs exceed
he Nyhol h eshold o associa ion wi h a diagnosis o p os-
a e cance , whe eas h ee SNPs in ALDH1A2 (in s ong LD
wi h each o he , he e o e ep esen ing one signal only)
exceed he Nyhol h eshold o associa ion wi h p os a e
cance -specific su i al. One SNP in ALDH1B1 also exceeds
he Nyhol h eshold o associa ion wi h p os a e cance -
specific su i al in low-g ade p os a e cance .
Gene ic a ian s in alcohol me abolising genes
The e is e idence ha gene ic a ia ions in ADH and ALDH
genes a ec ing e hanol me abolism
32–35
a e associa ed wi h
al e ed alcohol in ake and isk o alcohol dependence.
31,35–37
The mos ex ensi ely s udied SNP in he con ex o alcohol
in ake is s1229984 in ADH1B. I has been shown o be asso-
cia ed wi h inc eased ad e se e ec s om alcohol in ake and
educed consump ion.
35,36
On a e age mino allele ca ie s
d ink 17.2% ewe uni s/week (95%CI:15.6%, 18.9%), a e less
likely o be in he op hi d o alcohol d inking olume (odds
a io, OR 50.7; 95%CI:0.68,0.73) and a e less likely o binge
d ink (OR 50.78; 95%CI:0.73,0.84).
23
I has also been
epo ed o a ec cance isk a a ious si es.
5
In ou s udy,
he associa ions o his SNP wi h p os a e cance diagnosis
and su i al we e OR
fixed
51.00 (95%CI:0.96,1.03, p al-
ues 50.87), and HR
fixed
51.11 (95%CI:0.95,1.30, p al-
ues 50.17), espec i ely. Combining he e ec s o he
ADH1B SNP on alcohol in ake wi h he uppe confidence
in e als om ou esul s implies ha a 17% educ ion in
Table 2. Resul s o uni a ia e me a- eg essions o es i he associa ion o he wo SNPs ( ep esen ing he wo signals obse ed) is a ec ed
by selec ed s udy cha ac e is ics
Con idence in e als
Single nucleo ide polymo phism S udy cha ac e is ic Ra io o odds- a ios Lowe Uppe pValues
s1441817 PSA 1.48 0.75 2.93 0.19
(ALDH1A2) FHX 0.63 0.35 1.14 0.10
USA 0.69 0.30 1.57 0.30
Age 1.58 1.00 2.51 0.05
s10973794 PSA 0.65 0.23 1.80 0.33
(ALDH1B1) FHX 0.65 0.25 1.68 0.27
USA 1.30 0.29 5.72 0.65
Age 1.30 0.62 2.72 0.40
PSA, mean PSA a diagnosis.
FHX, pe cen age amily his o y.
USA, s udy loca ion in USA s. es o wo ld.
Age, mean age a diagnosis.
Cance Epidemiology
82 Alcohol me abolism and p os a e cance
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC
alcohol consump ion is unlikely o educe p os a e cance
isk by >3% and p os a e cance mo ali y by >5%.
Alcohol is me abolised o ace aldehyde, a known ca cino-
gen, and he e is e idence o suppo he heo y ha gene ic
a ian s in alcohol me abolising genes, which con ol he p o-
duc ion and b eakdown o ace aldehyde, con ibu e o ca ci-
nogenesis.
4,5,20,24
The e is also e idence o a issue-specific
in e ac ion in he p os a e be ween e hanol and e inoic acid,
h ough modula ions o ALDH1A1, ALDH1A2 and
ALDH1A3 le els.
38
To ou knowledge, his is he fi s com-
p ehensi e in es iga ion o he associa ion be ween ADH and
ALDH a ian s, as gene ic p oxies o alcohol, and p os a e
cance o da e. Gene ic p edisposi ion o p os a e cance has
been examined by GWASs, which shows common gene ic
a ian s can explain 33% he i abili y o p os a e cance bu
no genome-wide significan hi s a e in ADHs o ALDHs
39,40
(howe e , his lack o e idence om GWASs could be a ype
2 e o ). Simila ly, we did no find any e idence o gene ic
associa ion be ween ADH/ALDH a ian s and p os a e can-
ce incidence in his s udy. Possible easons o his include
ype 2 e o , especially i he unde lying e ec s o alcohol on
p os a e cance incidence a e small and limi ed o he e y
hea y d inking beha iou s and/o o he mo e agg essi e
o ms o disease, as possibly sugges ed by he ecen
li e a u e.
7,12
We ha e shown ha SNPs in ALDH1A2 a e associa ed
wi h al e ed p os a e cance -specific mo ali y in a case-only
analysis. None o hese SNPs appea o ha e egula o y ea-
u es (www.ensemble.o g), so hey a e unlikely o be causal
a ian s hemsel es bu a he hey could be in LD wi h he
causal a ian s. Recen ly, ALDH iso o ms ha e been sug-
ges ed as possible mechanis ic media o s o me as asis in
p os a e cance in pa icula
41
and o he solid umou s in
gene al.
42
One s udy ound lack o compelling e idence link-
ing a ia ion in ALDH1 (including ALDH1A1, ALDH1A2,
ALDH1A3 and ALDH1B1) wi h p os a e cance p og es-
sion,
41
bu ano he had epo ed p elimina y e idence o a
po en ial ole o ALDH1A2 as a umou supp esso gene in
p os a e cance cell lines
43
and dec eased exp ession o
ALDH1A2 has been associa ed wi h sho e ecu ence ee
su i al in pa ien s wi h p os a e cance .
43
In ou s udy, h ee
in onic SNPs in ALDH1A2 we e associa ed wi h longe su -
i al, none o which we e di ec ly o indi ec ly ( h ough LD)
associa ed wi h alcohol- ela ed pheno ypes (h p://www.ebi.ac.
uk/gwas/). One po en ial explana ion o ou esul s may be
ha hese SNPs, o o he s in LD wi h hem, lead o inc eased
ac i i y in ALDH1A2. We specula e ha he obse ed p os-
a e cance su i al e ec could be he esul o a ne inc ease
in he syn hesis o e inoic acid (by ALDH1A2), which is
pa icula ly beneficial when he a e o con e sion is a ec ed
by slowe ADH ac i i y in he p esence o alcohol consump-
ion ( e inol and e hanol bo h being ADH subs a es,
44
and
e hanol modula ing e inoic acid syn hesis in he a
p os a e
38
).
Ano he in onic SNP in ALDH1B1 (also close o
IGFBPL1) was ound o be associa ed wi h inc eased mo ali-
y ollowing a diagnosis o low-g ade p os a e cance .
ALDH1B1 is he second mos abundan mi ochond ial
ALDH, a e ALDH2, wi h documen ed in ol emen in alco-
hol me abolism and dependence.
45
Howe e , his specific
SNP is no known o be in LD wi h any o he a ian s asso-
cia ed wi h alcohol pheno ypes o da e, he e o e we canno
specula e on i s specific ole in ela ion o alcohol. E idence
has been p e iously ound linking le els o ALDH1B1 o su -
i al ollowing gas ic cance
46
and non-small-cell lung can-
ce ,
47
bu no p os a e cance ,
41
howe e he la e was a
s udy in i o Con e sely,ALDH1A1 exp ession in he p os-
a e has been epo ed o be a good candida e p ognos ic bio-
ma ke , based on all cause mo ali y and o a lesse ex en
p os a e cance -specific mo ali y,
48
and ALDH1A3 exp es-
sion is hough o be in ol ed wi h ini ia ion and p og ession
o se e al cance s,
49
howe e we did no obse e an associa-
ion wi h common ge m-line mu a ions in ei he o hese
genes, o did we obse e associa ions wi h a ia ion in ADH
genes. This could be due o di e en unc ional e ec s o a -
ian s on me abolic le els (e.g., alcohol and ace aldehyde peak
le els and cumula i e concen a ions), a ec ing p os a e can-
ce p oli e a ion o su i al di e en ly bo h in e ms o e ec
sizes and pa hways.
Di e ences by umou g ade
In his s udy, we ha e ound ha gene ic a ian s in genes
in ol ed in alcohol me abolism we e associa ed wi h disease-
specific mo ali y in men wi h p os a e cance , mos o
whom had been diagnosed wi h low-g ade disease. We ound
a signal specific o low-g ade p os a e cance su i al, bu
none o high-g ade disease. Possible easons why s onge
associa ions we e no seen wi h high-g ade cance s include:
limi ed powe , as he e we e smalle numbe s o high- com-
pa ed wi h low-g ade cases ( his was in es iga ed using a
e e se powe calcula ion, o assess he powe o his s udy o
de ec small associa ions (Supplemen a y ma e ial Table S3));
pa ien s beha iou may change ollowing a diagnosis, e.g.
pa ien s wi h high-g ade p os a e cance may be oo ill o
d ink and he e ec o he SNPs canno be seen in he
absence o d inking; o he findings could eflec a ue clini-
cal di e ence in he way alcohol a ec s su i al o he di -
e en g ades o p os a e cance .
S eng hs and limi a ions
The s eng hs o his s udy include he la ge sample size and
a ailabili y o da a on bo h isk and mo ali y s a ified by
g ade, which is an impo an p edic o o p ognosis. Impo -
an ly, we used a Mendelian andomisa ion app oach, which
minimises he po en ial o bias due o con ounding, in o -
ma ion bias ( ecall bias and sick-qui e e ec ) and e e se
causa ion, majo limi a ions o p e ious s udies in his
a ea.
7,8,13
We we e also able o con ol o con ounding by
popula ion s a ifica ion by adjus ing o basic popula ion
Cance Epidemiology
B unne e al. 83
In . J. Cance : 140, 75–85 (2017) V
C2016 The Au ho s In e na ional Jou nal o Cance published by John Wiley & Sons L d on behal o UICC