1
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
www.na u e.com/scien i ic epo s
Gene ic In e ac ions wi h Age,
Sex, Body Mass Index, and
Hype ension in Rela ion o A ial
Fib illa ion: The AFGen Conso ium
Lu-Chen Weng1,2, Ka h yn L. Lune a3,4, Ma ina Mülle -Nu asyid5,6,7, Albe Ve non Smi h8,9,
Sébas ien Thé iaul 10,11, Pe e E. Weeke12,13, John Ba na d 14, Joshua C. Bis15, Leo-Pekka
Lyy ikäinen 16, Ma cus E. Klebe 17, And eas Ma insson18, Hen y J. Lin19,20, Michiel
Riens a 21, S ella T ompe 22,23, Bouwe P. K ij he24, Ma cus Dö 25,26, De ek Kla in1,2,27,28,
Daniel I. Chasman29, Mo i z F. Sinne 5,7, Melanie Waldenbe ge 5,30,31, Leno e J. Laune 32,
Tama a B. Ha is32, Elsayed Z. Soliman33, Al a o Alonso34, Guillaume Pa é10,11, Ped o L.
Teixei a 35, Joshua C. Denny36, M. Benjamin Shoemake 37, Da id R. Van Wagone 38,
Jona han D. Smi h39, B uce M. Psa y40,41, Nona So oodehnia41,42, Ken D. Taylo 19,43, Mika
Kähönen44, Kjell Nikus45, G aciela E. Delgado17, Olle Melande 46,47, Gunna Engs öm46,
Jie Yao19,43, Xiuqing Guo19,43, Ing id E. Ch is ophe sen1,2,48, Pa ick T. Ellino 1,2, Bas iaan
Geelhoed21, Niek Ve weij 21, Pe e Mac a lane49, Ian Fo d50, Jan Hee inga24, Osca H.
F anco24, And é G. Ui e linden51, Uwe Völke 26,52, Alexande Teume 26,53, Lynda M. Rose54,
S e an Kääb5,7, Vilmundu Gudnason8,9, Dan E. A king55, Da id Conen56,57,58, Dan M. Roden59,
Mina K. Chung38,60, Susan R. Heckbe 41,61, Emelia J. Benjamin 3, Te ho Leh imäki16,
Win ied Mä z17,62, J. Gus a Smi h18, Je ome I. Ro e 19,63, Pim an de Ha s 21, J. Wou e
Jukema22,64,65, B uno H. S icke 66,67, S ephan B. Felix25,26, Ch is ine M. Albe 68 & S e en A.
Lubi z 1,2
I is unclea whe he gene ic ma ke s in e ac wi h isk ac o s o in luence a ial ib illa ion (AF) isk.
We pe o med genome-wide in e ac ion analyses be ween gene ic a ian s and age, sex, hype ension,
and body mass index in he AFGen Conso ium. S udy-speci ic esul s we e combined using me a-
analysis (88,383 indi iduals o Eu opean descen , including 7,292 wi h AF). Va ian s wi h nominal
in e ac ion associa ions in he disco e y analysis we e es ed o associa ion in ou independen
s udies (131,441 indi iduals, including 5,722 wi h AF). In he disco e y analysis, he AF isk associa ed
wi h he mino s6817105 allele (a he PITX2 locus) was g ea e among subjec s ≤ 65 yea s o age han
among hose > 65 yea s (in e ac ion p- alue = 4.0 × 10−5). The in e ac ion p- alue exceeded genome-
wide signi icance in combined disco e y and eplica ion analyses (in e ac ion p- alue = 1.7 × 10−8).
We obse ed one genome-wide signi ican in e ac ion wi h body mass index and se e al sugges i e
in e ac ions wi h age, sex, and body mass index in he disco e y analysis. Howe e , none was eplica ed
in he independen sample. Ou indings sugges ha he pa hogenesis o AF may di e acco ding
o age in indi iduals o Eu opean descen , bu we did no obse e e idence o s a is ically signi ican
gene ic in e ac ions wi h sex, body mass index, o hype ension on AF isk.
Recei ed: 27 Ap il 2017
Accep ed: 26 July 2017
Published: xx xx xxxx
OPEN
1Ca dio ascula Resea ch Cen e , Massachuse s Gene al Hospi al, Bos on, MA, USA. 2P og am in Medical and
Popula ion Gene ics, The B oad Ins i u e o MIT and Ha a d, Camb idge, MA, USA. 3Na ional Hea Lung and Blood
Ins i u e’s and Bos on Uni e si y’s F amingham Hea S udy, F amingham, MA, USA. 4Depa men o Bios a is ics,
Bos on Uni e si y School o Public Heal h, Bos on, MA, USA. 5DZHK (Ge man Cen e o Ca dio ascula Resea ch),
pa ne si e: Munich Hea Alliance, Munich, Ge many. 6Ins i u e o Gene ic Epidemiology, Helmhol z Zen um
München - Ge man Resea ch Cen e o En i onmen al Heal h, Neuhe be g, Ge many. 7Depa men o Medicine I,
Uni e si y Hospi al Munich, Ludwig-Maximilians-Uni e si y, Munich, Ge many. 8Icelandic Hea Associa ion, 201,
Kopa ogu , Iceland. 9Facul y o Medicine, Uni e si y o Iceland, 101, Reykja ik, Iceland. 10Popula ion Heal h Resea ch
www.na u e.com/scien i ic epo s/
2
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
A ial ib illa ion (AF) is a common a hy hmia and is associa ed wi h inc eased isk o s oke, hea ailu e, and
mo ali y1–4. P e ious s udies ha e demons a ed ha inc easing age, male sex, high blood p essu e, and obesi y
a e associa ed wi h highe AF isk5–11. AF is he i able12–17, and gene ic associa ion s udies ha e iden i ied 16 loci
agged by common gene ic a ian s ha a e associa ed wi h AF18–22.
Typically, genome-wide associa ion s udies ha e assumed ha he e ec o each es ed SNP on AF isk is con-
s an ac oss a ious isk ac o s, hough some da a sugges ha he e ec sizes may di e o di e en alues o
isk ac o s. Fo example, a ian s a he HIATL1 egion ha e been shown o in e ac wi h alcohol consump ion
o a ec colo ec al cance isk23. Unde s anding he di e ences in magni udes o e ec o SNPs in ela ion o
Ins i u e, Hamil on, Canada. 11Depa men o Pa hology and Molecula Medicine, McMas e Uni e si y, Hamil on,
Canada. 12Depa men o Medicine, Vande bil Uni e si y, Nash ille, TN, USA. 13Depa men o Ca diology, The
Hea Cen e, Rigshospi ale , Copenhagen, Denma k. 14Depa men o Quan i a i e Heal h Sciences, Le ne
Resea ch Ins i u e, Cle eland Clinic, Cle eland, OH, USA. 15Ca dio ascula Heal h Resea ch Uni , Depa men o
Medicine, Uni e si y o Washing on, Sea le, WA, USA. 16Depa men o Clinical Chemis y, Fimlab Labo a o ies and
Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e, Tampe e, Finland. 17V h Depa men o Medicine
(Neph ology, Hype ensiology, Endoc inology, Diabe ology, Rheuma ology), Medical Facul y o Mannheim,
Uni e si y o Heidelbe g, Mannheim, Ge many. 18Depa men o Ca diology, Lund Uni e si y and Skåne Uni e si y
Hospi al, Lund, Sweden. 19Ins i u e o T ansla ional Genomics and Popula ion Sciences and Depa men o
Pedia ics, Los Angeles Biomedical Resea ch Ins i u e a Ha bo -UCLA Medical Cen e , To ance, Cali o nia, USA.
20Di ision o Medical Gene ics, Depa men o Pedia ics, Ha bo -UCLA Medical Cen e , To ance, Cali o nia, USA.
21Depa men o Ca diology, Uni e si y o G oningen, Uni e si y Medical Cen e G oningen, G oningen, The
Ne he lands. 22Depa men o Ca diology, Leiden Uni e si y Medical Cen e , Leiden, The Ne he lands. 23Depa men
o Ge on ology and Ge ia ics, Leiden Uni e si y Medical Cen e , Leiden, The Ne he lands. 24Depa men o
Epidemiology, E asmus Medical Cen e , Uni e si y Medical Cen e , Ro e dam, The Ne he lands. 25Depa men o
In e nal Medicine B, Uni e si y Medicine G ei swald, G ei swald, Ge many. 26DZHK (Ge man Cen e o
Ca dio ascula Resea ch), pa ne si e G ei swald, G ei swald, Ge many. 27Cen e o Human Gene ic Resea ch,
Ca dio ascula Resea ch Cen e , Massachuse s Gene al Hospi al, Ha a d Medical School, Bos on, MA, USA.
28Depa men o Su ge y, Massachuse s Gene al Hospi al, Bos on, MA, USA. 29Di isions o P e en i e Medicine and
Gene ics, B igham and Women’s Hospi al, Bos on, MA, USA. 30Resea ch uni o Molecula Epidemiology, Helmhol z
Zen um München - Ge man Resea ch Cen e o En i onmen al Heal h, Neuhe be g, Ge many. 31Ins i u e o
Epidemiology II, Helmhol z Zen um München - Ge man Resea ch Cen e o En i onmen al Heal h, Neuhe be g,
Ge many. 32Labo a o y o Epidemiology and Popula ion Sciences, Na ional Ins i u e on Aging, In amu al Resea ch
P og am, Na ional Ins i u es o Heal h, Be hesda, Ma yland, 20892, USA. 33Epidemiological Ca diology Resea ch
Cen e (EPICARE), Wake Fo es School o Medicine, Wins on Salem, NC, USA. 34Depa men o Epidemiology,
Rollins School o Public Heal h, Emo y Uni e si y, A lan a, GA, USA. 35Depa men o Biomedical In o ma ics,
Vande bil Uni e si y, Nash ille, TN, USA. 36Depa men s o Medicine and Biomedical In o ma ics, Vande bil
Uni e si y, Nash ille, TN, USA. 37Depa men o Medicine, Vande bil Uni e si y Medical Cen e , Nash ille, TN, USA.
38Depa men o Molecula Ca diology, Le ne Resea ch Ins i u e, Cle eland Clinic, Cle eland, OH, USA.
39Depa men o Cellula and Molecula Medicine, Le ne Resea ch Ins i u e, Cle eland Clinic, Cle eland, OH, USA.
40Ca dio ascula Heal h Resea ch Uni , Depa men s o Medicine, Epidemiology and Heal h Se ices, Uni e si y o
Washing on, Sea le, WA, USA. 41Kaise Pe manen e Washing on Heal h Resea ch Ins i u e, Kaise Founda ion
Heal h Plan o Washing on, Sea le, WA, USA. 42Ca dio ascula Heal h Resea ch Uni , Di ision o Ca diology,
Depa men s o Medicine and Epidemiology, Uni e si y o Washing on, Sea le, WA, USA. 43Di ision o Genomic
Ou comes, Depa men o Pedia ics, Ha bo -UCLA Medical Cen e , To ance, Cali o nia, USA. 44Depa men o
Clinical Physiology, Tampe e Uni e si y Hospi al and Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e,
Tampe e, Finland. 45Depa men o Ca diology, Hea Cen e , Tampe e Uni e si y Hospi al and Facul y o Medicine
and Li e Sciences, Uni e si y o Tampe e, Tampe e, Finland. 46Depa men o Clinical Sciences, Lund Uni e si y,
Malmö, Sweden. 47Depa men o In e nal Medicine, Skåne Uni e si y Hospi al, Malmö, Sweden. 48Depa men o
Medical Resea ch, Bæ um Hospi al, Ves e Viken Hospi al T us , Sand ika, No way. 49Ins i u e o Heal h and
Wellbeing, Uni e si y o Glasgow, Sco land, Uni ed Kingdom. 50Robe son Cen e o Bios a is ics, Uni e si y o
Glasgow, Sco land, Uni ed Kingdom. 51Depa men o In e nal Medicine, E asmus Uni e si y Medical Cen e ,
Ro e dam, The Ne he lands. 52In e acul y Ins i u e o Gene ics and Func ional Genomics, Uni e si y Medicine
G ei swald, G ei swald, Ge many. 53Ins i u e o Communi y Medicine, Uni e si y Medicine G ei swald, G ei swald,
Ge many. 54Di ision o P e en i e Medicine, B igham and Women’s Hospi al, Bos on, MA, USA. 55McKusick-Na hans
Ins i u e o Gene ic Medicine, Johns Hopkins Uni e si y School o Medicine, Bal imo e, MD, USA. 56Popula ion
Heal h Resea ch Ins i u e, McMas e Uni e si y, Hamil on, Canada. 57Uni e si y Hospi al Basel, Basel, Swi ze land.
58Ca dio ascula Resea ch Ins i u e Basel, Basel, Swi ze land. 59Depa men s o Medicine, Clinical Pha macology,
and Biomedical In o ma ics, Vande bil Uni e si y, Nash ille, TN, USA. 60Hea and Vascula Ins i u e, Cle eland
Clinic, Cle eland, OH, USA. 61Ca dio ascula Heal h Resea ch Uni and Depa men o Epidemiology, Uni e si y o
Washing on, Sea le, WA, USA. 62Clinical Ins i u e o Medical and Chemical Labo a o y Diagnos ics, Medical
Uni e si y G az, G az, Aus ia. 63Di ision o Genomic Ou comes, Depa men s o Pedia ics and Medicine, Ha bo -
UCLA Medical Cen e , To ance, Cali o nia, USA. 64Ein ho en Labo a o y o Expe imen al Vascula Medicine, LUMC,
Leiden, The Ne he lands. 65In e uni e si y Ca diology Ins i u e o he Ne he lands, U ech , The Ne he lands.
66Depa men o Epidemiology and In e nal Medicine, E asmus Uni e si y Medical Cen e Ro e dam, U ech , The
Ne he lands. 67Inspec o a e o Heal h Ca e, U ech , The Ne he lands. 68Di isions o P e en i e Medicine and
Ca dio ascula Medicine, B igham and Women’s Hospi al, Bos on, MA, USA. Co espondence and eques s o
ma e ials should be add essed o S.A.L. (email: [email p o ec ed])
www.na u e.com/scien i ic epo s/
3
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
AF ac oss common clinical isk ac o s could po en ially e ine ou knowledge abou he gene ic basis o AF in
impo an clinical subse s o indi iduals. Ne e heless, no la ge sys ema ic examina ion o in e ac ions be ween
gene ic a ian s and clinical AF isk ac o s has been conduc ed.
We he e o e aimed o de e mine whe he common gene ic a ian s in e ac wi h age, sex, hype ension, and
body mass index o modi y AF isk in a la ge sample o indi iduals o Eu opean ances y.
Resul s
A o al o 88,378 subjec s, including 7,292 wi h AF, we e included in he disco e y analysis (Table1). The numbe s
o included SNPs and alues o genomic in la ion ac o s (λ) o each s udy (a e applying quali y con ol c i e ia
o SNP exclusions) a e displayed in Supplemen al Table1. O e all, genomic in la ion ac o s anged om 0.85 o
1.2 ac oss s udies and in e ac ion analyses. Quan ile-quan ile (QQ) plo s o expec ed e sus obse ed in e ac ion
p- alue dis ibu ions o associa ions o he app oxima ely 2.5 million au osomal SNPs o each in e ac ion anal-
ysis a e displayed in Supplemen al Fig.1a–d. Manha an plo s o -log10 (p- alue) agains he physical coo dina es
o he 22 au osomes a e shown in Fig.1A–D.
In e ac ions wi h isk ac o s a known AF loci. We i s e alua ed he associa ions be ween gene ic
in e ac ions and clinical ac o s (age, sex, hype ension, and body mass index) wi h AF a 16 es ablished AF
suscep ibili y loci om p io genome-wide associa ion s udies (Supplemen al Table2; signi icance h esh-
old = 6.25 × 10-4, see me hods o explana ion). We obse ed signi ican in e ac ions wi h age o SNP s6817105
(ups eam o PITX2 a ch omosome locus 4q25; in e ac ion p- alue = 4 × 10-5; Table2). The mino C allele o
SNP s6817105 was associa ed wi h a g ea e isk o AF among indi iduals 65 yea s o age o younge [odds a io
(OR) = 1.75, 95% CI 1.61–1.91, p = 6.2 × 10-36], han among pa icipan s olde han 65 yea s (OR = 1.38, 95% CI
1.28–1.47, p = 6.3 × 10-17). Among o he known AF loci, SNP s3807989 a he CAV1 locus displayed a nominal
in e ac ion wi h age ha was no s a is ically signi ican (in e ac ion p = 2.9 × 10-3; Table2). Howe e , he majo
G allele was associa ed wi h highe AF isk in he younge g oup (OR = 1.25, 95% CI 1.16–1.34, p = 3.6 × 10-10
o subjec s ≤ 65 yea s; OR = 1.09, 95% CI 1.03–1.15, p = 1.4 × 10-3 o subjec s >65). We did no obse e any
signi ican in e ac ions be ween AF-associa ed SNPs and sex, hype ension, o body mass index.
N wi h
AF N o al Males, n (%) Age,
mean ± SD Hype ension,
n (%) Body mass index,
kg/m2, mean ± SD
Disco e y s udies
Inciden AF
AGES*158 2718 1011(37.2) 76.3 ± 5.46 2144 (78.9) 76.27 ± 5.46
ARIC*799 9053 4255 (47.0) 54.3 ± 5.7 2426 (26.8) 27.0 ± 4.8
CHS*763 3185 1234 (38.7) 72.2 ± 5.3 1680 (52.8) 26.3 ± 4.4
FHS*306 4025 1751 (43.5) 64.7 ± 12.6 1988 (49.5) 27.7 ± 5.2
MESA*155 2526 1206 (47.74) 62.66 ± 10.24 975 (38.6) 27.74 ± 5.06
PREVEND*113 3520 1811 (50) 49.5 ± 12.4 1157 (30) 26.1 ± 4.3
PROSPER*505 5244 2524 (48.1) 75.34 ± 3.35 3257 (62.1) 26.82 ± 4.18
RS*591 5665 2282 (40.3) 69.1 ± 8.98 3081 (54.4) 26.32 ± 3.69
WGHS*648 20842 0 (0) 54.6 ± 7.0 5022 (24) 25.3 ± 6.7
P e alen AF
AFNET/KORA 448 886 524 (59.1) 53.4 ± 7.8 326 (36.8) 27.9 ± 4.6
AGES 241 2959 1154 (39.0) 76.47 ± 5.50 2359 (79.8) 27.06 ± 4.44
BioVU o1 238 4766 2552 (53.6) 62.2 ± 16.3 3270 (68.6) 26.2 ± 11.2
BioVU 660 120 3790 1722 (45.4) 62.8 ± 15.9 1966 (51.9) 24.0 ± 15.0
CCAF 807 2661 1918 (72.1) 61.7 ± 11.15 1793 (67.4) 29.5 ± 5.78
FHS 253 4401 1957 (44.5) 65.4 ± 12.8 2215 (50.5) 27.70 ± 5.16
LURIC 361 2959 2077 (70.2) 63.0 ± 10.6 2154 (72.8) 27.5 ± 4.02
MGH/MIGEN 366 1277 780 (61.1) 49.5 ± 9.7 — —
RS 309 5974 2427 (40.6) 69.4 ± 9.1 3273 (54.8) 26.3 ± 3.69
SHIP 107 1923 927 (48.2) 50.97 ± 15.07 496 (25.8) 27.33 ± 4.56
Replica ion s udies
Inciden AF
MDCS 876 7353 3800 (48) 58.8 ± 6.6 5010 (68) 26.1 ± 4.1
P e alen AF
BEAT-AF 1520 3040 1795 (59) 51.7 ± 18.6 1363 (45) 25.8 ± 4.4
FINCAVAS 940 3021 1835 (61) 61.9 ± 14 2117 (70) 27.5 ± 4.5
UK Biobank 2386 118027 55669 (47) 56.9 ± 7.9 25307 (21) 27.5 ± 4.8
Table 1. Subjec Cha ac e is ics. Abb e ia ions: AF: a ial ib illa ion; NA: no a ailable; SD: s anda d
de ia ion. *In o ma ion a DNA collec ion.
www.na u e.com/scien i ic epo s/
4
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
In e ac ions wi h isk ac o s in genome-wide analyses. Table3 displays he esul s o SNP in e -
ac ions wi h AF isk ac o s ac oss he genome. The mos signi ican gene ic in e ac ion ha exceeded ou
genome-wide signi icance h eshold (an in e ac ion p- alue < 4 × 10-8, see me hods o explana ion) was
obse ed o SNP s12416673 wi h body mass index (in e ac ion p = 2.9 × 10-8; 6.4 kb ups eam o COL13A1 a
ch omosome egion 10q21; Table3; Supplemen al Figu e2). Speci ically, wi h each 1-uni inc ease in body mass
index, each copy o he mino A allele o SNP s12416673 was associa ed wi h an inc eased isk o AF (in e ac-
ion β = 0.0224, in e ac ion p = 2.9 × 10-8). Addi ionally, we obse ed 8 loci ha exhibi ed sugges i e in e ac ions
wi h AF isk ac o s (i.e., he in e ac ion p- alue was < 1 × 10-6 o he op SNP, and wo o mo e SNPs in he
Figu e 1. Manha an plo s o gene ic in e ac ions wi h age, sex, body mass index, and hype ension in ela ion
o AF isk. The ed line shows he signi ican in e ac ion p- alue h eshold (p < 4 × 10-8), and he blue line
shows he sugges i e signi ican in e ac ion p- alue h eshold (p < 1 × 10-6).
www.na u e.com/scien i ic epo s/
5
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
same egion exhibi ed in e ac ion p- alues < 1 × 10-5). Speci ically, we obse ed in e ac ions wi h age a 2 loci,
sex a 1 locus, and body mass index a 5 loci (Table3). No gene ic in e ac ions wi h hype ension exceeded he
sugges i e genome-wide o adjus ed AF suscep ibili y locus signi icance h esholds.
SNP A1/
A2 A1
eq Loc Closes
gene
SNP and AF isk ac o in e ac ion
Age Sex Body mass index Hype ension
In e ac ion β *(se) p In e ac ion β (se) pIn e ac ion β (se) pIn e ac ion β (se) p
s6666258 C/G 0.30 1q21 KCNN3 0.0979 (0.048) 0.04 0.0092 (0.045) 0.84 -0.0023 (0.005) 0.62 0.024 (0.045) 0.60
s3903239 G/A 0.44 1q24 PRRX1 0.0661 (0.044) 0.13 -0.050 (0.041) 0.23 0.0021 (0.004) 0.63 -0.014 (0.041) 0.74
s4642101 G/T 0.65 3p25 CAND2 0.0828 (0.047) 0.08 0.0425 (0.045) 0.35 -0.0024 (0.005) 0.59 0.0813 (0.045) 0.07
s1448818 C/A 0.25 4q25 PITX2 0.0207 (0.049) 0.67 0.0285 (0.046) 0.54 0.0094 (0.005) 0.04 -0.0597 (0.045) 0.19
s6817105 C/T 0.13 4q25 PITX2 0.2420 (0.059) 4.0 × 10-50.0065 (0.055) 0.91 0.0078 (0.006) 0.16 -0.0516 (0.055) 0.35
s4400058 A/G 0.09 4q25 PITX2 0.0665 (0.070) 0.34 -0.0343 (0.068) 0.61 -0.0051 (0.007) 0.47 -0.0406 (0.066) 0.54
s6838973 C/T 0.57 4q25 PITX2 0.0636 (0.045) 0.16 -0.0599 (0.043) 0.16 -0.0005 (0.004) 0.90 -0.0823 (0.042) 0.05
s13216675 T/C 0.69 6q22 GJA1 -0.0287 (0.050) 0.57 0.0869 (0.047) 0.07 0.0064 (0.005) 0.18 0.0616 (0.046) 0.18
s3807989 G/A 0.60 7q31 CAV1 0.1329 (0.045) 2.9 × 10-3-0.0054 (0.041) 0.90 -0.0003 (0.004) 0.95 -0.0603 (0.041) 0.14
s10821415 A/C 0.42 9q22 C9o 3 0.0736 (0.047) 0.11 -0.0039 (0.044) 0.93 0.0012 (0.004) 0.79 -0.0813 (0.043) 0.06
s10824026 A/G 0.84 10q22 SYNPO2L -0.0035 (0.063) 0.96 -0.0213 (0.059) 0.72 0.0139 (0.006) 0.02 0.258 (0.060) 0.66
s12415501 T/C 0.16 10q24 NEURL 0.0701 (0.064) 0.27 0.0994 (0.058) 0.09 0.0011 (0.006) 0.85 0.0684 (0.057) 0.23
s10507248 T/G 0.73 12q24 TBX5 -0.0573 (0.050) 0.25 0.0571 (0.046) 0.22 -0.0025 (0.005) 0.59 -0.0208 (0.046) 0.65
s1152591 A/G 0.48 14q23 SYNE2 0.0178 (0.045) 0.69 -0.0082 (0.042) 0.85 0.004 (0.004) 0.35 0.0255 (0.042) 0.54
s7164883 G/A 0.16 15q24 HCN4 -0.0294 (0.058) 0.61 0.0284 (0.054) 0.60 0.0016 (0.006) 0.78 -0.0271 (0.055) 0.62
s2106261 T/C 0.18 16q22 ZFHX3 0.0106 (0.057) 0.85 0.0434 (0.054) 0.42 -0.0021 (0.006) 0.71 0.110 (0.053) 0.04
Table 2. Mul iplica i e SNP in e ac ions wi h AF isk ac o s a known AF loci. The signi icance h eshold
0.01/16 = 6.25 × 10-4. Abb e ia ions: AF: a ial ib illa ion; A1: allele 1; he isk allele was de ined based on a
p io GWAS56; A2: allele 2; A1 eq: allele 1 equency; Loc: locus; p: P- alue o he in e ac ion be ween he
isk ac o and he SNP. *In e ac ion β was om eg ession using an addi i e model. In e ac ion β (se) was
calcula ed as he me a-analysis log(e ec ) in subjec s ≤ 65 yea s o age minus he me a-analysis log(e ec )
in subjec s >65 yea s o age, o as he mul iplica i e in e ac ion be ween SNP* isk ac o o sex ( emales s.
males), hype ension (hype ensi e s. no ), and body mass index (pe 1 uni inc emen ).
SNP Loc Closes gene A1/
A2
Disco e y Replica ion Combined
A1
eq
(%) In e ac ion β (se) p
A1
eq
(%) In e ac ion β
(se)*p
A1
eq
(%) In e ac ion β
(se)*p
SNP x Age
s6817105†4q25 PITX2 C/T 0.13 0.2420 (0.059) 4.0 × 10-50.11 0.2213 (0.067) 9.5 × 10-40.12 0.2420 (0.043) 1.7 × 10-8
s3807989†7q31 CAV1 G/A 0.60 0.1329 (0.045) 2.9 × 10-30.59 -0.0531 (0.050) 0.28 0.59 0.0325 (0.032) 3.1 × 10-1
s2356251 14q22 MAP4K5 C/G 0.06 0.5716 (0.109) 1.6 × 10-70.04 -0.1894 (0.123) 0.12 0.05 0.2294 (0.081) 4.5 × 10-3
s1572779 20q13 MIR548AG2 G/T 0.10 0.3468 (0.070) 7.9 × 10-70.09 0.1456 (0.090) 0.10 0.10 0.2446 (0.054) 5.7 × 10-6
SNP x Sex
s2730668 12q21 TRHDE T/C 0.76 0.2734 (0.052) 1.7 × 10-70.76 0.0846 (0.0563) 0.13 0.76 0.1860 (0.0383) 1.2 × 10-6
SNP x Body Mass Index
s9394492 6q21 BTBD9 T/C 0.36 0.0222 (0.004) 2.7 × 10-70.38 -0.0070 (0.005) 0.15 0.37 0.0092 (0.003) 4.1 × 10-3
s1874425 8q21 AD A1A T/C 0.25 0.0231 (0.005) 9.3 × 10-70.25 0.0070 (0.005) 0.18 0.25 0.0160 (0.004) 5.4 × 10-6
s1545567 9p24 VLDLR T/C 0.64 -0.0256 (0.005) 4.3 × 10-70.65 -0.0010 (0.005) 0.85 0.65 -0.0131 (0.004) 2.3 × 10-4
s12416673 10q21 COL13A1 A/G 0.43 0.0224 (0.004) 2.9 × 10-80.43 -0.0018 (0.005) 0.71 0.43 0.0122 (0.003) 7.1 × 10-5
s6062828 20q13 LOC105372719/YTHDF1 C/G 0.68 -0.0245 (0.005) 8.6 × 10-70.69 -0.0105 (0.005) 0.03 0.69 -0.0174 (0.004) 6.5 × 10-7
Table 3. Disco e y and eplica ion analysis esul s o op SNP in e ac ions wi h AF isk ac o s. Abb e ia ions:
AF: a ial ib illa ion; A1: allele 1; he isk allele was de ined based on a p io GWAS56; A2: allele 2; A1 eq:
allele 1 equency; Loc: locus; p: P- alue o he in e ac ion be ween he isk ac o and he SNP. *In e ac ion
β was om eg ession using addi i e model. In e ac ion β (se) was calcula ed as he me a-analysis log(e ec )
in subjec s ≤ 65 yea s o age minus he me a-analysis log(e ec ) in subjec s > 65 yea s o age, o as he
mul iplica i e in e ac ion be ween SNP* isk ac o o sex ( emales s. males) and body mass index (pe 1 uni
inc emen ). †Known AF loci.

www.na u e.com/scien i ic epo s/
6
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
Replica ion. In o al, we selec ed 10 SNP in e ac ions (Table3) o eplica ion associa ion es ing in ou
independen coho s (131,441 indi iduals, including 5,722 wi h AF). Only one in e ac ion emained signi ican ly
associa ed wi h AF. SNP s6817105 a he 4q25 locus exhibi ed a signi ican in e ac ion wi h age (in e ac ion
p = 9.5 × 10-4). As in ou disco e y analysis, among indi iduals wi h he mino C allele o s6817105, hose ≤65
yea s old had a g ea e isk o AF (OR = 1.80; 95% CI 1.67–1.95, p = 6.6 × 10-52), han pa icipan s olde han 65
yea s (OR = 1.45; 95% CI 1.30–1.61, p = 1.4 × 10-11). Simila ly, s6817105 was associa ed wi h a 27% highe AF
isk in subjec s ≤65 yea s o age (compa ed wi h subjec s >65 yea s o age) in he combined disco e y and epli-
ca ion analysis (in e ac ion p = 1.7 × 10-8; Fig.2). A g ea e isk o AF o he s6817105 C allele was obse ed in
pa icipan s aged 65 yea s o younge (OR = 1.78; 95% CI 1.68–1.89, p = 5.6 × 10-86) han in pa icipan s olde
han 65 yea s (OR = 1.40; 95% CI 1.32–1.49, p = 7.8 × 10-27).
Powe calcula ion. Gi en he lack o obse ed associa ions be ween SNP in e ac ions wi h clinical isk ac-
o s and AF, we pe o med powe calcula ions o es ima e powe o disco e y using Quan o24 (h p://bios a s.usc.
edu/Quan o.h ml; Fig.3). As an example, we es ima ed powe o obse e a SNP in e ac ion wi h sex, assuming
a popula ion comp ised o 50% males, an AF popula ion p e alence o 1%, and a case o con ol a io o 1:10 (as
in ou s udy). We modeled a main e ec OR o 1.5 o sex, and a gene ic odds a io o 1.5 o a SNP. We es ima ed
ha >100,000 AF cases would be a equi ed o achie e 80% powe o such an e ec size, indica ing ha we had
limi ed powe o de ec all bu subs an ial gene ic in e ac ions wi h clinical isk ac o s.
Discussion
In ou analysis o ~88,000 indi iduals o Eu opean ances y, including 7,292 indi iduals wi h AF, we obse ed
ha he well-es ablished AF locus a ch omosome egion 4q25 ( agged by s6817105) was associa ed wi h a di -
e en ial isk o AF acco ding o age. Speci ically, he OR o each copy o he mino s6817105 allele was 1.78
o indi iduals ≤65 yea s o age, compa ed o 1.40 o indi iduals >65 yea s. Beyond he age in e ac ion wi h he
4q25 locus, we did no obse e any signi ican in e ac ions be ween gene ic a ian s and age, sex, hype ension,
o body mass index a e eplica ion a emp s. These indings sugges ha s ong gene ic in e ac ions wi h he AF
isk ac o s s udied in his manusc ip a e unlikely o be p ominen mechanisms d i ing AF suscep ibili y.
Ou indings suppo and ex end p io s udies examining gene ic in e ac ions o AF. Fo example, op a -
ian s a he 4q25 ch omosome locus, ups eam o PITX2, we e associa ed wi h g ea e AF isks among younge
indi iduals in seconda y analyses o a genome-wide associa ion s udy20. Howe e , no o mal s a is ical es
Figu e 2. Age-s a i ied associa ion be ween he ch omosome 4q25 locus and AF in he combined da ase o
p ima y and eplica ion s udies. OR and Pmain e e o he odds a io and p- alue o he associa ion es be ween
s6817105 and AF isk in each age-s a um. Pin e ac ion e e s o he p- alue co esponding o he di e ence in
e ec sizes be ween he wo age s a a es ed.
Figu e 3. Numbe o cases equi ed o de ec in e ac ion odds a ios be ween 1.01 o 1.5 wi h common SNPs
(mino allele equencies (MAF) o 0.05–0.5) wi h 80% powe assuming an AF p e alence o 1%, 50% males,
SNP ma ginal e ec odds a ios o 1.5, sex ma ginal e ec odds a ios o 1.5, case:con ol a ios o 1:10, and
α = 4 × 10-8. Powe calcula ions we e pe o med using Quan o24.
www.na u e.com/scien i ic epo s/
7
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
o in e ac ion was pe o med. G ea e e ec sizes o o he AF suscep ibili y SNPs a he 4q25 locus we e also
obse ed among younge a he han olde indi iduals in some, bu no all, coho s in a la ge eplica ion s udy25.
Mo eo e , in keeping wi h ou obse a ions, p io s udies did no ind e idence ha AF isk is modi ied by in e -
ac ions be ween SNPs a he 4q25 locus and sex20. Ou indings demons a e ha gene ic a ia ion a he 4q25
locus is, on a e age, associa ed wi h g ea e isks o ea ly-onse AF.
Ou indings a e consis en wi h epidemiologic obse a ions demons a ing g ea e he i abili y o ea lie
onse o AF16. The s onge e ec o he 4q25 locus on AF in a younge popula ion implies ha he con ibu ion
o his locus o AF suscep ibili y may be mo e ele an o hose wi h ea ly-onse AF, a he han la e onse o ms.
O e all, ou obse a ion ha gene ic a ia ion a he 4q25 locus is associa ed wi h AF (beyond genome-wide
signi ican h esholds) in bo h younge and olde indi iduals unde sco es he p edominan ole o his locus in
AF pa hogenesis– ega dless o age.
PITX2 is a homeobox ansc ip ion ac o in ol ed in speci ica ion o pulmona y ascula u e26, ca diac la -
e ali y27, and supp ession o a le a ial sinoa ial-node like pacemake 28. He e ozygous null Pi x2c mouse hea s
a e mo e suscep ible o pacing induced AF han a e wild- ype coun e pa s29. The ela i e oles o PITX2 egula-
ion in AF suscep ibili y in bo h human de elopmen and in adul li e a e unclea . Fu u e la ge s udies a e wa -
an ed o sys ema ically de e mine whe he he e a e di e en age-speci ic e iologic sub ypes o AF, and whe he
PITX2 modula ion a ies wi h age acco ding o geno ype.
Al hough ou analysis was no designed o speci ically quan i y he con ibu ion o gene ic ac o s o AF he -
i abili y, he absence o obse ed in e ac ions be ween AF and sex, body mass index, and hype ension sugges s
ha common a ian in e ac ions wi h hese clinical isk ac o s a e unlikely o explain a subs an ial p opo ion
o a iance in AF suscep ibili y. La ge s udies will be necessa y o accu a ely quan i y he con ibu ions o bo h
common and a e a ia ion, epigene ic mechanisms, copy-numbe a ia ion, epis a ic e ec s, and o he en i on-
men al in e ac ions ha may in luence AF he i abili y. Mo eo e , u he examina ion is needed o de e mine
he ex en o which he 4q25 locus, he p edominan suscep ibili y locus o AF, explains he he i abili y o he
condi ion.
Ou s udy should be in e p e ed in he con ex o he s udy design. Fi s , we included indi iduals o Eu opean
ances y only, so ou inding may no be gene alizable o o he acial g oups. Second, AF isk ac o s we e a ail-
able only a he ime o AF onse in case-con ol s udies, a he han be o e AF onse , po en ially biasing owa d
he null any biologically ele an SNP by isk ac o in e ac ions ha may occu yea s be o e he onse o AF.
Howe e , we suspec ha such misclassi ica ion o isk ac o s a us is unlikely o ha e esul ed in sys ema ic bias
o body mass index (which ends o be ela i ely s able o e ime30) and age (because an in e ac ion wi h age
and he PITX2 locus is suppo ed by p io obse a ions). Thi d, ou sample size p o ided limi ed powe o iden-
i y in e ac ions wi h ela i ely small e ec sizes. Addi ionally, he use o mo e powe ul s a is ical app oaches31,
non-mul iplica i e in e ac ions, and inclusion o addi ional AF isk ac o s, may acili a e iden i ica ion o loci a
which gene ic in e ac ions exis in ela ion o AF. Fou h, ou single SNP in e ac ion app oach does no exclude
a lack o in e ac ion wi h polygenic suscep ibili y o AF. Fi h, we acknowledge ha AF may be clinically un ec-
ognized, leading o misclassi ica ion o AF s a us, and ha we lacked powe o analyze AF sub ypes sepa a ely.
Fu u e analyses wi h addi ional a hy hmia ou comes may help cla i y he ole o gene ic in e ac ions wi h isk
ac o s ac oss a ange o a hy hmia pheno ypes.
In summa y, we iden i ied a signi ican in e ac ion wi h age a he AF suscep ibili y locus on ch omosome
4q25 ups eam o PITX2 in indi iduals o Eu opean ances y. Despi e se e al sugges i e SNP in e ac ions wi h
common AF isk ac o s in disco e y analyses, we did no obse e subs an ial e idence o such in e ac ions as
common mechanisms unde lying AF isk.
Me hods
S udy popula ion. Disco e y coho s included he: Ge man Compe ence Ne wo k o A ial Fib illa ion and
Coope a i e Heal h Resea ch in he Region Augsbu g (AFNET/KORA); Age, Gene/En i onmen Suscep ibili y
Reykja ik S udy (AGES) s udy; A he oscle osis Risk in Communi ies (ARIC) s udy; Vande bil elec onic med-
ical eco d-linked DNA eposi o y (BioVU); Cle eland Clinic Lone AF s udy (CCAF); Ca dio ascula Heal h
S udy (CHS); F amingham Hea S udy (FHS); Ludwigsha en Risk and Ca dio ascula Heal h (LURIC) s udy;
Mul i-E hnic S udy o A he oscle osis (MESA); Massachuse s Gene al Hospi al Lone AF s udy and Myoca dial
In a c ion Gene ics Conso ium (MGH/MIGEN); P e en ion o Renal and Vascula End-s age Disease
(PREVEND) s udy; PROspec i e S udy o P a as a in in he Elde ly a Risk (PROSPER); Ro e dam S udy
(RS); S udy o Heal h in Pome ania (SHIP); and Women’s Genome Heal h S udy (WGHS). Replica ion s ud-
ies included he: Basel A ial Fib illa ion Coho S udy (Bea -AF); Finnish Ca dio ascula S udy (FINCAVAS);
Malmo die and cance s udy (MDCS); and UK Biobank. De ailed desc ip ions o each s udy ha e been p e i-
ously epo ed (Supplemen al Me hods and Supplemen al Table3).
The s udy p o ocol was app o ed by he E hical Commi ee/ins i u ional e iew boa ds o Ludwig Maximilian
Uni e si y o Munich, Na ional Bioe hics Commi ee, Johns Hopkins Bloombe g School o Public Heal h,
Uni e si y o Minneso a, Vande bil Uni e si y Medical Cen e , Cle eland Clinic, Uni e si y o Washing on,
Bos on Uni e si y Medical Campus, Rhineland-Pala ina e S a e Chambe o Physicians, Massachuse s Gene al
Hospi al, Uni e si y Medical Cen e G oningen, Leiden Uni e si y Medical Cen e , E asmus MC - Uni e si y
Medical Cen e Ro e dam, Uni e si y Medicine G ei swald, B igham and Women’s Hospi al, e hics commi ee
no hwes /cen al Swi ze land, e hics commi ee Zu ich, Pi kanmaa Hospi al Dis ic , and Lund Uni e si y. All
MESA s udy si es ecei ed app o al o conduc his esea ch om local ins i u ional e iew boa ds a : Columbia
Uni e si y ( o he MESA New Yo k Field Cen e ), Johns Hopkins Uni e si y ( o he MESA Bal imo e Field
Cen e ), No hwes e n Uni e si y ( o he MESA Chicago Field Cen e ), Uni e si y o Cali o nia, Los Angeles
( o he MESA Los Angeles Field Cen e ), Uni e si y o Minneso a ( o he MESA Twin Ci ies Field Cen e ),
Wake Fo es Uni e si y Heal h Sciences Cen e ( o he MESA Wins on-Salem Field Cen e ). W i en in o med
www.na u e.com/scien i ic epo s/
8
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
consen was ob ained om all s udy subjec s o hei p oxies (excep BioVU, which is a de-iden i ied EMR bio e-
posi o y and was “op -ou ” p io o Decembe 2014). All expe imen s we e pe o med in acco dance wi h ele-
an guidelines and egula ions.
AF asce ainmen . Asce ainmen o AF and isk ac o s in each s udy has been desc ibed p e iously10, 14, 32–54,
De ailed desc ip ions a e p o ided in Supplemen al Table3. We de ined p e alen AF as an e en ha was diag-
nosed a o p io o an indi idual’s DNA collec ion in coho s udies and on he basis o AF asce ainmen in
case-con ol s udies. We de ined inciden AF as an e en ha was diagnosed a e DNA collec ion among pa ici-
pan s ee o clinically appa en AF a DNA collec ion in coho s udies. All AF isk ac o s excep age we e asce -
ained a he ime o DNA collec ion. Age was de ined a DNA collec ion o a he da e o ec ui men in coho
s udies, and a ime o AF diagnosis ( o AF cases) o a ime o DNA collec ion ( o con ols) in case-con ol
s udies.
Exposu e asce ainmen . Sex was de ined on he basis o sel - epo . Pa icipan s we e classi ied as ha ing
hype ension i he sys olic blood p essu e was ≥140 mm Hg o he dias olic blood p essu e was ≥90 mm Hg
a any clinic isi o exam an eceden o DNA collec ion, o i he pa icipan was ecei ing ea men wi h an
an ihype ensi e medica ion and had a sel - epo ed his o y o hype ension o high blood p essu e a he ime
o DNA collec ion (no applicable in ARIC o FHS; Supplemen al Table3). Body mass index was de ined as he
weigh (kg) di ided by he heigh (m) squa ed. Blood p essu e measu emen s, medica ion lis s, weigh s, and
heigh s we e asce ained acco ding o s udy-speci ic p o ocols. All pa icipan s in he disco e y analysis we e
geno yped on genome-wide SNP a ay pla o ms (Supplemen al Table4). Impu ed geno ypes used in ou analysis
included app oxima ely 2.5 million gene ic a ian s om he HapMap CEU sample ( elease 22).
S a is ical analysis. Fo each indi idual s udy, logis ic eg ession ( o p e alen AF; o inciden AF in
MESA and PREVEND only), gene alized es ima ing equa ions (in FHS o accoun o ela ed indi iduals), o
Cox p opo ional haza d eg ession ( o inciden AF in p ospec i e coho s o he han MESA and PREVEND)
we e pe o med o examine whe he AF was associa ed wi h in e ac ions be ween SNP and AF isk ac-
o s. Fo Cox models, pe son- ime began a s udy baseline, and indi iduals we e censo ed a dea h o loss o
ollow-up. Robus a iance es ima es we e used when easible. De ails o he eg ession models a e desc ibed in
Supplemen al Table4. All models we e adjus ed o age (age a baseline o inciden AF, and age a AF onse o
p e alen AF), sex, si e (ARIC and CHS), sub-coho (FHS), s udy-speci ic co a ia es, and popula ion s uc u e,
i applicable. SNPs wi h low impu a ion quali y (R-squa e < 0.3) o a mino allele equency < 0.05 we e emo ed
om he analysis.
Fo in e ac ion analyses in ol ing sex, hype ension, and body mass index, main e ec e ms o each isk ac-
o , as well as mul iplica i e in e ac ion e ms be ween each SNP and he espec i e isk ac o , we e included in
he eg ession models. Rega ding analyses o age, nonlinea associa ions be ween SNPs and age could po en ially
go unde ec ed, due o a iable dis ibu ions o age ac oss he s udies in ou analysis. Addi ionally, some s udies
had only o mos ly ea ly-onse /la e-onse AF cases, which limi ed ou abili y o pe o m a eg ession model wi h
dicho omized age in such samples. The e o e, we assessed SNP in e ac ions wi h age by compa ing me a-analysis
es ima es o associa ions be ween each SNP and AF in indi iduals ≤65 e sus >65 yea s o age (see below).
S udies wi h <100 AF e en s in each s a um o age we e no included, in o de o a oid uns able e ec es ima es.
Es ima o s o mul iplica i e in e ac ion e ms we e me a-analyzed o sex, hype ension, and body mass
index analyses in METAL55, using an in e se- a iance weigh ed ixed-e ec s app oach wi h genomic-con ol
co ec ion. Fo age, we pe o med an in e se- a iance weigh ed ixed-e ec s me a-analysis o he es ima o s o
each SNP sepa a ely wi hin each age s a um, wi h genomic-con ol co ec ion. Es ima o s we e compa ed using
a Z es , as men ioned abo e.
SNPs wi h absolu e e ec sizes ≥3 o SNPs ha we e a ailable in only one s udy we e excluded om ou
inal esul s, o minimize he likelihood o spu ious alse posi i e indings. Fo each o he ou genome-wide
in e ac ion assessmen s, we employed an expe imen -wide wo sided alpha h eshold o 0.05, which we adjus ed
o mul iple hypo hesis es ing. We dis ibu ed he alpha di e en ially ac oss he genome, acco ding o a p io i
hypo heses abou in e ac ions be ween SNPs and AF isk ac o s. Speci ically, we dis ibu ed one- i h o he alpha
o each o he 16 mos signi ican ly associa ed SNPs a genome-wide signi ican loci iden i ied in p io s udies56, 57
(in e ac ion p < 0.01/16 = 6.25 × 10-4). The emaining ou - i hs o he alpha we e dis ibu ed e enly ac oss he
genome, o an alpha h eshold o 4 × 10-8 (in e ac ion p < 0.04/~1,000,000 independen es s).
Signi ican ly associa ed SNPs and SNPs wi h sugges i e associa ions (i.e., an in e ac ion p < 0.005 a a ecog-
nized AF GWAS locus; o an in e ac ion p < 1 × 10-6 combined wi h in e ac ion p < 1 × 10-5 o wo addi ional
SNPs wi hin he same ±50 kb egion) in he disco e y analysis we e ca ied o wa d o eplica ion es ing. In
o al, we ca ied o wa d 10 SNPs o eplica ion es ing (see below), and he e o e assumed a eplica ion in e ac-
ion p h eshold o 0.005 (0.05/10 SNPs). The esul s o eplica ion s udies alone, as well as combined wi h esul s
om disco e y s udies, we e me a-analyzed as desc ibed abo e.
Re e ences
1. Ch is iansen, C. B. e al. A ial ib illa ion and isk o s oke: a na ionwide coho s udy. Eu opace: Eu opean pacing, a hy hmias, and
ca diac elec ophysiology: jou nal o he wo king g oups on ca diac pacing, a hy hmias, and ca diac cellula elec ophysiology o he
Eu opean Socie y o Ca diology, eu 401 (2016).
2. Ve mond, R. A. e al. Incidence o a ial ib illa ion and ela ionship wi h ca dio ascula e en s, hea ailu e, and mo ali y: a
communi y-based s udy om he Ne he lands. Jou nal o he Ame ican College o Ca diology 66, 1000–1007 (2015).
3. A onow, W. S. & Banach, M. A ial Fib illa ion: The New Epidemic o he Ageing Wo ld. Jou nal o a ial ib illa ion 1, 154,
doi:10.4022/ja ib.154 (2009).
www.na u e.com/scien i ic epo s/
9
Scien i ic RepoR s | 7: 11303 | DOI:10.1038/s41598-017-09396-7
4. Sha i , N. e al. Ra e-con ol e sus hy hm-con ol s a egies and ou comes in sep uagena ians wi h a ial ib illa ion. The Ame ican
jou nal o medicine 126, 887–893, doi:10.1016/j.amjmed.2013.04.021 (2013).
5. Benjamin, E. J. e al. Independen isk ac o s o a ial ib illa ion in a popula ion-based coho : he F amingham Hea S udy. Jama
271, 840–844 (1994).
6. Iguchi, Y. e al. P e alence o A ial Fib illa ion in Communi y-Dwelling Japanese Aged 40 Yea s o Olde in Japan Analysis o
41,436 Non-Employee Residen s in Ku ashiki-Ci y. Ci cula ion Jou nal 72, 909–913 (2008).
7. Kannel, W. B. & Benjamin, E. J. S a us o he epidemiology o a ial ib illa ion. Medical Clinics o No h Ame ica 92, 17–40 (2008).
8. Kannel, W. B., Wol , P. A., Benjamin, E. J. & Le y, D. P e alence, incidence, p ognosis, and p edisposing condi ions o a ial
ib illa ion: popula ion-based es ima es. The Ame ican jou nal o ca diology 82, 2N–9N (1998).
9. K ahn, A. D., Man eda, J., Ta e, R. B., Ma hewson, F. A. L. & Cuddy, T. E. The na u al his o y o a ial ib illa ion: incidence, isk
ac o s, and p ognosis in he Mani oba Follow-Up S udy. The Ame ican jou nal o medicine 98, 476–484 (1995).
10. Smi h, J. G., Pla ono , P. G., Hedblad, B., Engs om, G. & Melande , O. A ial ib illa ion in he Malmo Die and Cance s udy: a
s udy o occu ence, isk ac o s and diagnos ic alidi y. Eu opean jou nal o epidemiology 25, 95–102, doi:10.1007/s10654-009-
9404-1 (2010).
11. Wanahi a, N. e al. A ial ib illa ion and obesi y— esul s o a me a-analysis. Ame ican hea jou nal 155, 310–315 (2008).
12. A na , D. O. e al. Familial agg ega ion o a ial ib illa ion in Iceland. Eu opean hea jou nal 27, 708–712 (2006).
13. Ch is ophe sen, I. E. e al. Familial agg ega ion o a ial ib illa ion a s udy in Danish wins. Ci cula ion: A hy hmia and
Elec ophysiology 2, 378–383 (2009).
14. Ellino , P. T., Yoe ge , D. M., Ruskin, J. N. & MacRae, C. A. Familial agg ega ion in lone a ial ib illa ion. Human gene ics 118,
179–184 (2005).
15. Fox, C. S. e al. Pa en al a ial ib illa ion as a isk ac o o a ial ib illa ion in o sp ing. Jama 291, 2851–2855 (2004).
16. Lubi z, S. A. e al. Associa ion be ween amilial a ial ib illa ion and isk o new-onse a ial ib illa ion. Jama 304, 2263–2269
(2010).
17. Ma cus, G. M. e al. A i s -deg ee amily his o y in lone a ial ib illa ion pa ien s. Hea hy hm: he o icial jou nal o he Hea
Rhy hm Socie y 5, 826–830 (2008).
18. Benjamin, E. J. e al. Va ian s in ZFHX3 a e associa ed wi h a ial ib illa ion in indi iduals o Eu opean ances y. Na u e gene ics 41,
879–881, doi:10.1038/ng.416 (2009).
19. Ellino , P. T. e al. Common a ian s in KCNN3 a e associa ed wi h lone a ial ib illa ion. Na u e gene ics 42, 240–244 (2010).
20. Gudbja sson, D. F. e al. Va ian s con e ing isk o a ial ib illa ion on ch omosome 4q25. Na u e 448, 353–357 (2007).
21. Gudbja sson, D. F. e al. A sequence a ian in ZFHX3 on 16q22 associa es wi h a ial ib illa ion and ischemic s oke. Na u e
gene ics 41, 876–878 (2009).
22. Sinne , M. F. e al. In eg a ing gene ic, ansc ip ional, and unc ional analyses o iden i y i e no el genes o a ial ib illa ion.
Ci cula ion, CIRCULATIONAHA-114 (2014).
23. Gong, J. e al. Genome-Wide In e ac ion Analyses be ween Gene ic Va ian s and Alcohol Consump ion and Smoking o Risk o
Colo ec al Cance . PLoS gene ics 12, e1006296 (2016).
24. Gaude man, W. J. Sample size equi emen s o ma ched case‐con ol s udies o gene–en i onmen in e ac ion. S a is ics in medicine
21, 35–50 (2002).
25. Kääb, S. e al. La ge scale eplica ion and me a-analysis o a ian s on ch omosome 4q25 associa ed wi h a ial ib illa ion. Eu opean
hea jou nal 30, 813–819 (2009).
26. Momme s eeg, M. T. M. e al. Pi x2c and Nkx2-5 a e equi ed o he o ma ion and iden i y o he pulmona y myoca dium.
Ci cula ion esea ch 101, 902–909 (2007).
27. Tessa i, A. e al. Myoca dial Pi x2 di e en ially egula es he le a ial iden i y and en icula asymme ic emodeling p og ams.
Ci cula ion esea ch 102, 813–822 (2008).
28. Wang, J. e al. Pi x2 p e en s suscep ibili y o a ial a hy hmias by inhibi ing le -sided pacemake speci ica ion. P oceedings o he
Na ional Academy o Sciences 107, 9753–9758 (2010).
29. Ki chho , P. e al. PITX2c is exp essed in he adul le a ium, and educing Pi x2c exp ession p omo es a ial ib illa ion inducibili y
and complex changes in gene exp ession. Ci cula ion: Ca dio ascula Gene ics 4, 123–133 (2011).
30. Heo, M., Fai h, M. S. & Pie obelli, A. Resis ance o change o adul hood body mass index. In e na ional jou nal o obesi y and ela ed
me abolic diso de s: jou nal o he In e na ional Associa ion o he S udy o Obesi y 26, 1404–1405 (2002).
31. Manning, A. K. e al. Me a-analysis o gene-en i onmen in e ac ion: join es ima ion o SNP and SNP x en i onmen eg ession
coe icien s. Gene ic epidemiology 35, 11–18, doi:10.1002/gepi.20546 (2011).
32. Alonso, A. e al. Incidence o a ial ib illa ion in whi es and A ican-Ame icans: he A he oscle osis Risk in Communi ies (ARIC)
s udy. Ame ican hea jou nal 158, 111–117 (2009).
33. Ambale-Venka esh, B. e al. Dias olic unc ion assessed om agged MRI p edic s hea ailu e and a ial ib illa ion o e an 8-yea
ollow-up pe iod: he mul i-e hnic s udy o a he oscle osis. Eu opean Hea Jou nal-Ca dio ascula Imaging 15, 442–449 (2014).
34. Bild, D. E. e al. Mul i-e hnic s udy o a he oscle osis: objec i es and design. Ame ican jou nal o epidemiology 156, 871–881 (2002).
35. Conen, D. e al. Alcohol consump ion and isk o inciden a ial ib illa ion in women. Jama 300, 2489–2496 (2008).
36. Dawbe , T. R., Meado s, G. F. & Moo e, F. E. J . Epidemiological App oaches o Hea Disease: The F amingham S udy*. Ame ican
Jou nal o Public Heal h and he Na ions Heal h 41, 279–286 (1951).
37. F ied, L. P. e al. The ca dio ascula heal h s udy: design and a ionale. Annals o epidemiology 1, 263–276 (1991).
38. Ha is, T. B. e al. Age, Gene/En i onmen Suscep ibili y–Reykja ik S udy: mul idisciplina y applied phenomics. Ame ican jou nal
o epidemiology 165, 1076–1087 (2007).
39. Hee inga, J. e al. P e alence, incidence and li e ime isk o a ial ib illa ion: he Ro e dam s udy. Eu opean hea jou nal 27,
949–953 (2006).
40. Ho man, A. e al. The Ro e dam S udy: objec i es and design upda e. Eu opean jou nal o epidemiology 22, 819–829 (2007).
41. John, U. e al. S udy o Heal h In Pome ania (SHIP): a heal h examina ion su ey in an eas Ge man egion: objec i es and design.
Sozial-und P ä en i medizin 46, 186–194 (2001).
42. Kannel, W. B., Feinleib, M., McNama a, P. M., Ga ison, R. J. & Cas elli, W. P. An in es iga ion o co ona y hea disease in amilies
The F amingham o sp ing s udy. Ame ican jou nal o epidemiology 110, 281–290 (1979).
43. Ka hi esan, S. e al. Genome-wide associa ion o ea ly-onse myoca dial in a c ion wi h single nucleo ide polymo phisms and copy
numbe a ian s. Na u e gene ics 41, 334–341 (2009).
44. Nabaue , M. e al. The Regis y o he Ge man Compe ence NETwo k on A ial Fib illa ion: pa ien cha ac e is ics and ini ial
managemen . Eu opace: Eu opean pacing, a hy hmias, and ca diac elec ophysiology: jou nal o he wo king g oups on ca diac pacing,
a hy hmias, and ca diac cellula elec ophysiology o he Eu opean Socie y o Ca diology 11, 423–434 (2009).
45. Nieminen, T. e al. The Finnish Ca dio ascula S udy (FINCAVAS): cha ac e ising pa ien s wi h high isk o ca dio ascula
mo bidi y and mo ali y. BMC ca dio ascula diso de s 6, 9 (2006).
46. Rex ode, K. M., Lee, I. M., Cook, N. R., Hennekens, C. H. & Bu ing, J. E. Baseline cha ac e is ics o pa icipan s in he Women’s
Heal h S udy. Jou nal o women’s heal h & gende -based medicine 9, 19–27 (2000).
47. Ridke , P. M. e al. Ra ionale, design, and me hodology o he Women’s Genome Heal h S udy: a genome-wide associa ion s udy o
mo e han 25 000 ini ially heal hy Ame ican women. Clinical chemis y 54, 249–255 (2008).