ARTICLE OPEN
Common gene ic a ian s con ibu e o he i abili y o age a
onse o schizoph enia
Es e Sada-Fuen e
1,2
, Selena A anda
1,2
, Se gi Papiol
3,4
, U s Heilb onne
3
, Ma ía Dolo es Mol ó
2,5,6
, Edua do J. Aguila
2,6,7,8
,
Ja ie González-Peñas
2,9
, Ál a o And eu-Be nabeu
2,9,10
, Celso A ango
2,9,10
, Benedic o C espo-Faco o
2,11,12,13
,
Ana González-Pin o
2,14
, Lou des Fañanás
2,15
, Ba ba a A ias
2,15
, Julio Bobes
2,16,17,18
, Ja ie Cos as
19
, Lou des Ma o ell
1,2
,
Thomas G. Schulze
3,4,20
, Janos L. Kalman
3,4
, Elisabe Vilella
1,2
✉and Ge a d Mun ané
1,2,21
✉
© The Au ho (s) 2023, co ec ed publica ion 2023
Schizoph enia (SCZ) is a complex diso de ha ypically a ises in la e adolescence o ea ly adul hood. Age a onse (AAO) o SCZ is
associa ed wi h long- e m ou comes o he disease. We explo ed he gene ic a chi ec u e o AAO wi h a genome-wide associa ion
s udy (GWAS), he i abili y, polygenic isk sco e (PRS), and copy numbe a ian (CNV) analyses in 4 740 subjec s o Eu opean
ances y. Al hough no genome-wide significan locus was iden ified, SNP-based he i abili y o AAO was es ima ed o be be ween
17 and 21%, indica ing a mode a e con ibu ion o common a ian s. We also pe o med c oss- ai PRS analyses wi h a se o
men al diso de s and iden ified a nega i e associa ion be ween AAO and common a ian s o SCZ, childhood mal ea men and
a en ion-defici /hype ac i i y diso de . We also in es iga ed he ole o copy numbe a ian s (CNVs) in AAO and ound an
associa ion wi h he leng h and numbe o dele ions (P- alue =0.03), whe eas he p esence o CNVs p e iously epo ed in SCZ was
no associa ed wi h ea lie onse . To ou knowledge, his is he la ges GWAS o AAO o SCZ o da e in indi iduals om Eu opean
ances y, and he fi s s udy o de e mine he in ol emen o common a ian s in he he i abili y o AAO. Finally, we e idenced he
ole played by highe SCZ load in de e mining AAO bu disca ded he ole o pa hogenic CNVs. Al oge he , hese esul s shed ligh
on he gene ic a chi ec u e o AAO, which needs o be confi med wi h la ge s udies.
T ansla ional Psychia y (2023) 13:201 ; h ps://doi.o g/10.1038/s41398-023-02508-0
INTRODUCTION
Schizoph enia (SCZ) is a complex diso de influenced by an
in ica e in e play o gene ic and en i onmen al ac o s. SCZ
pa ien s show subs an ial he e ogenei y in clinical cha ac e is ics
such as symp oma ology, cogni i e abili y, cou se, o e all
unc ioning, and age a onse (AAO). AAO has been consis en ly
included among he mos impo an de e minan s o disease
ou come and is widely accep ed as a significan clinical and
p ognos ic ac o [1,2]. Fo ins ance, an ea lie AAO is associa ed
wi h a highe likelihood o ha ing ela i es wi h SCZ [3,4], and has
been co ela ed wi h an inc eased numbe o hospi aliza ions and
illness episodes, mo e equen nega i e symp oms, and poo e
cogni ion, o e all unc ioning, and global ou come [5–7]. In
gene al, men ha e an ea lie AAO, usually be ween 20 and 24
yea s o age, while in women, he onse occu s be ween 25 and 35
yea s o age [8–10]. In addi ion, women appea o ha e a
seconda y peak a ound menopause, be ween 50 and 54 yea s old
[7,11].
The gene ic a chi ec u e o SCZ is complex. He i abili y
es ima es om win and popula ion-based s udies ange be ween
64 and 81% [12,13], in which common gene ic a ian s accoun
o a la ge p opo ion (24.4%) [14]. The e is also e idence ha
Recei ed: 17 Janua y 2023 Re ised: 23 May 2023 Accep ed: 2 June 2023
1
Hospi al Uni e si a i Ins i u Pe e Ma a, Ins i u d’In es igació Sani à ia Pe e Vi gili (IISPV), Depa men o Psychia y, Uni e si a Ro i a i Vi gili (URV), Reus, Spain.
2
Cen o de
In es igación Biomédica en Red de Salud Men al (CIBERSAM), Mad id, Spain.
3
Ins i u e o Psychia ic Phenomics and Genomics (IPPG), Uni e si y Hospi al, Ludwig Maximilian
Uni e si y o Munich, 80336 Munich, Ge many.
4
Depa men o Psychia y and Psycho he apy, Uni e si y Hospi al, Ludwig Maximilian Uni e si y o Munich, 80336 Munich,
Ge many.
5
Depa men o Gene ics, Uni e si a de Valencia, 46100 Valencia, Spain.
6
Biomedical Resea ch Ins i u e INCLIVA, 46010 Valencia, Spain.
7
Depa men o Psychia y,
Hospi al Clínico Uni e si a io de Valencia, 46010 Valencia, Spain.
8
Facul y o Medicine, Uni e sidad de Valencia, 46010 Valencia, Spain.
9
Depa men o Child and Adolescen
Psychia y, Ins i u e o Psychia y and Men al Heal h, G ego io Ma añón Heal h Resea ch Ins i u e (IiSGM), Hospi al Gene al Uni e si a io G ego io Ma añón, 28007 Mad id, Spain.
10
Facul y o Medicine, Uni e sidad Complu ense, 28007 Mad id, Spain.
11
Depa men o Psychia y, Uni e sidad de Can ab ia, 39005 San ande , Can ab ia, Spain.
12
Hospi al
Uni e si a io Ma qués de Valdecilla-IDIVAL, 39008 San ande , Can ab ia, Spain.
13
Depa men o Psychia y, Uni e si y Hospi al Vi gen del Rocío, Ins i u o de Biomedicina de
Se illa (IBiS), Se illa, Spain.
14
Depa men o Psychia y, Hospi al Uni e si a io A aba, Ins i u o de In es igación Sani a ia Bioa aba, Uni e sidad del País Vasco, 01009 Vi o ia, Spain.
15
Depa men o E olu iona y Biology, Ecology and En i onmen al Sciences, Facul y o Biology, Uni e si a de Ba celona, Ins i u de Biomedicina de la Uni e si a de Ba celona
(IBUB), 08028 Ba celona, Spain.
16
Facul y o Medicine and Heal h Sciences - Psychia y, Uni e sidad de O iedo, 33006 O iedo, Spain.
17
Men al Heal h Se ices o P incipado de
As u ias (SESPA), Ins i u o de In es igación Sani a ia del P incipado de As u ias (ISPA), 33011 O iedo, Spain.
18
Ins i u o de Neu ociencias del P incipado de As u ias (INEUROPA),
33003 O iedo, Spain.
19
Psychia ic Gene ics G oup, Ins i u o de In es igación Sani a ia de San iago de Compos ela (IDIS), Se izo Galego de Saúde (SERGAS), Complexo
Hospi ala io Uni e si a io de San iago de Compos ela (CHUS), 15706 San iago de Compos ela, Spain.
20
Depa men o Psychia y and Beha io al Sciences, SUNY Ups a e Medical
Uni e si y, Sy acuse, NY, US.
21
Ins i u de Biologia E olu i a (UPF-CSIC), Depa amen de Medicina i Ciències de la Vida, Uni e si a Pompeu Fab a, Pa c de Rece ca Biomèdica de
Ba celona, Ba celona, Spain. ✉email: ilellae@pe ema a.com; [email p o ec ed]om
www.na u e.com/ p
T ansla ional Psychia y
1234567890();,:
bo h a e single-nucleo ide a ian s and a e copy numbe
a ian s (CNVs) con ibu e o he isk o de eloping SCZ [15–19].
In ac , indi iduals wi h a pa hogenic CNV ep esen mo e han 2%
o he confi med cases [20]. CNVs a e highly pene an and may
cause ea ly-onse o ms o de elopmen al delay o au ism
spec um diso de s. Thus, simila ly, i has been sugges ed ha
he p esence o CNVs may play an impo an ole in he onse o
SCZ, al hough hei con ibu ion is s ill unclea [20].
The he i abili y o he AAO has been es ima ed in sibling pai s a
33% [21], indica ing a mode a e gene ic basis. Howe e , in
con as wi h he as amoun o in o ma ion on he gene ics o
SCZ ob ained om genome-wide associa ion s udies (GWAS)
[14,22,23], he gene ic de e minan s unde lying AAO emain
la gely unknown. To da e, only h ee GWAS ha e been pe o med
in ela i ely small coho s (<3000 indi iduals) and none o hem
iden ified any genomic loci associa ed wi h AAO a genome-wide
significance [24–26]. On he o he hand, ecen GWAS ca ied ou
based on he age a onse o bo h Bipola Diso de (BD) and Majo
Dep ession Diso de (MDD) wi h la ge sample sizes ha e
de e mined a significan SNP-based he i abili y and sha ed
gene ic isk wi h o he psychia ic diso de s [27,28]. Thus, u he
s udies wi h la ge sample sizes a e equi ed o es ima e he
con ibu ion o common gene ic a ian s o AAO and he
he i abili y hey may explain.
Ul ima ely, iden i ying and esea ching he gene ic ac o s ha
influence he AAO o SCZ may imp o e ou unde s anding o he
de elopmen and p og ession o his disease, p o ide new a ge s
o he apy, and acili a e he de elopmen o pe sonalized
he apeu ic in e en ions and p e en i e measu es. In his s udy,
we aimed o explo e he gene ic a chi ec u e o AAO by
pe o ming (1) a GWAS me a-analysis o nea ly 5000 subjec s o
Eu opean ances y, (2) he i abili y es ima es based on common
gene ic a ian s, (3) polygenic isk sco e (PRS) analyses wi h a se
o men al ai s, and finally 4) an assessmen o he influence o
known CNVs on AAO.
MATERIAL AND METHODS
Sample
Fou di e en da ase s, wo om Eu ope (CIBERSAM, PsyCou se) and wo
om USA (GAIN [29], and nonGAIN), we e ob ained and combined o
pe o m a GWAS me a-analysis on AAO comp ising 4740 pa ien s o
Eu opean ances y (Table 1). In all da ase s, subjec s me he a ia o SCZ,
schizoa ec i e diso de , schizoph eni o m diso de , delusional diso de , b ie
psycho ic diso de , o psycho ic diso de no o he wise specified, in he
Diagnos ic and S a is ical Manual o Men al Diso de s e sion IV (DSM-IV).
The CIBERSAM and PsyCou se da ase s we e collec ed by he au ho s,
while GAIN and nonGAIN da ase s we e ob ained om public esou ces.
Pa icipan s in he CIBERSAM (Biomedical Resea ch Ne wo k in Men al
Heal h) da ase we e ec ui ed om psychia ic in-pa ien uni s a se en
di e en hospi als in Spain [30]. Pa icipa ion was app o ed by he e hical
commi ees a he hospi als in ol ed in he ec ui men . Finally, samples
we e geno yped using he Illumina Infinium PsychA ay a he B oad
Ins i u e as pa o he wa e 3 me a-analysis GWAS o SCZ o he
Psychia ic Genomics Conso ium (PGC-SCZ wa e 3) [31]. The PsyCou se
samples we e pa o a mul i-si e Ge man/Aus ian longi udinal s udy
(www.psycou se.de) ha was conduc ed be ween 1 Janua y 2012 and 31
Decembe 2019. The s udy collec ed deep pheno ypic, neu opsychologic,
and omics da a om pa ien s wi h b ie psycho ic diso de , majo
dep essi e diso de (MDD), bipola diso de (BD), SCZ, schizoa ec i e
diso de , and heal hy indi iduals. Adul pa icipan s we e e e ed by he
clinical s a o iden ified by que ying pa ien egis ies. S udy p o ocols
we e e iewed and app o ed by he e hics commi ees o he Medical
Cen e s and Facul ies in ol ed in he ec ui men , in acco dance wi h he
Decla a ion o Helsinki. All pa icipan s p o ided w i en in o med consen .
The pheno ype in o ma ion was ga he ed using he 4.1 e sion o he
PsyCou se da a elease. These samples we e geno yped using he Illumina
Infinium PsychA ay [32]. Finally, GAIN and nonGAIN da ase s we e bo h
ob ained om he dbGaP eposi o y (accession numbe s phs000021. 3.p2
and phs000167. 1.p1, espec i ely), and geno yped wi h he A yme ix
Genome-Wide Human SNP A ay 6.0 as desc ibed elsewhe e [33].
Table 1. Desc ip ion o samples used in he s udy.
Da ase Sou ce N% Females Mean AAO
(SD)
Geno yping chip
CIBERSAM Se en g oups om he Biomedical Resea ch Ne wo k in Men al Heal h (CIBERSAM) 1704 32.98 25.32 (8.84) Illumina Infinium PsychA ay
PsyCou se The PsyCou se s udy 499 38.88 26.26 (9.55) Illumina Infinium PsychA ay
GAIN The Genome-Wide Associa ion S udy o Schizoph enia (dbGaP eposi o y s udy accession:
phs000021. 3.p2)
1280 29.92 21.11 (6.77) A yme ix Genome-Wide Human SNP
A ay 6.0
nonGAIN The Molecula Gene ics o Schizoph enia—nonGAIN Sample (MGS_nonGAIN, dbGaP
eposi o y s udy accession: phs000167. 1.p1)
1224 31.54 21.77 (7.24) A yme ix Genome-Wide Human SNP
A ay 6.0
E. Sada-Fuen e e al.
2
T ansla ional Psychia y (2023) 13:201
Age a onse
In he CIBERSAM da ase , AAO was defined as he onse o he fi s
psycho ic symp oms. The pa ien (and/o amily membe s) and he
psychia is defined when he fi s psychological symp oms appea ed.
In cases whe e his in o ma ion was no a ailable, we used he da e o
he fi s psychia ic isi due o a psycho ic episode. In he PsyCou se
da ase , AAO was collec ed as bo h he age a fi s ou pa ien and
inpa ien ea men . Subjec s wi h in o ma ion a ailable o ei he o
hese da a we e included, and when bo h da a we e a ailable, he
ea lies age was used. Fo he GAIN and nonGAIN da ase s, AAO was
defined as he mos likely AAO o psycho ic symp oms consis en wi h
he onse o SCZ. A consensus diagnos ician (PI o senio esea ch
clinician delega e) e iewed he diagnos ic a ings made independen ly
by wo esea ch diagnos icians (one o which could be he consensus
diagnos ician as well) and assigned a final diagnosis and AAO i he
a ings we e in ag eemen . The Kolmogo o -Smi no es was used o
de e mine pai -wise di e ences in AAO be ween da ase s.
GWAS and unc ional analyses
Quali y con ol (QC) was conduc ed o each da ase sepa a ely (CIBERSAM,
PsyCou se, GAIN and nonGAIN) using PLINK 1.9 [34], acco ding o s anda d
p ocedu es o GWAS [35]. B iefly, gene ic a ian s wi h missingness a e
>2%, mino allele equency (MAF) < 5%, Ha dy-Weinbe g equilib ium
(HWE) P- alue < 1e−06, and hose belonging o non-au osomal ch omo-
somes we e excluded om downs eam analyses. Ambiguous and
mul iallelic a ian s we e also emo ed. Subjec s wi h a missingness a e
> 2%, inc eased o dec eased he e ozygosi y a es (defined as ±3 s anda d
de ia ions away om he sample mean), and ela edness >12.5%
(PI_HAT > 0.125) we e excluded. Sex was impu ed based on X ch omo-
some he e ozygosi y/homozygosi y a es, be o e emo ing sex ch omo-
somes. P incipal componen analyses (PCA) we e conduc ed using
SMARTPCA om EIGENSOFT 6.1.4 [36]. To keep only subjec s wi h
Eu opean ances y, we did no use hose indi iduals who we e beyond
±3 s anda d de ia ions om he mean o he fi s wo p incipal
componen s (PCs) o he Eu opean clus e o he 1000 Genomes P ojec
[37] Phase I. We also emo ed ou indi iduals who clus e ed in he Finnish
subg oup o he Eu opean clus e . Be o e geno ype impu a ion, a PCA was
pe o med again on he emaining subjec s and he op 10 PCs we e kep
o u he analysis.
Geno ype impu a ion was conduc ed o each esul ing da ase
independen ly using he TOPMed e e ence panel [38]. Impu ed
da ase s we e fil e ed acco ding o an impu a ion quali y sco e ( -
squa ed) <0.9 and con e ed o bina y files using PLINK’s-- c flag.
Then, a pos -impu a ion QC was conduc ed o each da ase sepa a ely,
using PLINK. Only single-nucleo ide polymo phisms (SNPs) we e kep
o u he analyses. In addi ion, ambiguous and mul iallelic SNPs, as
well as SNPs wi h MAF < 1%, and an HWE P- alue < 1e−06 we e
excluded. The esul ing da ase s we e li ed o e o genome build 19
using he UCSC li O e Plink ool [39]. The CIBERSAM da ase included:
1704 subjec s and 4,962,031 SNPs; PsyCou se: 499 subjec s and
5,338,835 SNPs; GAIN: 1278 subjec s and 6,042,664 SNPs; and nonGAIN:
1259 subjec s and 6,074,765 SNPs. A GWAS was pe o med by linea
eg ession o each da ase in PLINK, using no malized AAO as ou come
and sex and he op 10 PCs as co a ia es. Since AAO was di e en
be ween da a se s a me a-analysis was hen conduc ed using he ool
METAL [40] applying an in e se a iance s a egy. As a esul , we
ob ained in o ma ion on 4740 subjec s and 6,540,522 SNPs. As an
al e na i e me hod, indi idual-le el impu ed geno ype da a o each o
he ou sepa a e da ase s was me ged and a GWAS was conduc ed in
pa allel o compa ison pu poses. F om he e on, his app oach is called
he me ged app oach (Supplemen a y No e).
Genomic loci showing sugges i e associa ions wi h AAO (P- alue < 1e-
05) we e iden ifiedandexplo edusing heFUMAso wa e[41]. Each
genomic isk locus was ep esen ed by he op lead SNP which had he
minimum P- alue in he locus. Lead SNPs we e defined as associa ed
SNPs (P- alue < 1e−05) ha we e independen o each o he a LD
2
< 0.1. Independen significan SNPs we e definedasSNPswi haP-
alue < 1e−05, and independen o each o he a a linkage disequili-
b ium (LD) h eshold
2
< 0.6. The genomic isk loci we e mapped o
p o ein-coding genes by posi ional mapping based on ANNOVAR [42]
and eQTL mapping wi h GTEx 8 and BRAINEAC da abases. Finally,
pa hway en ichmen analyses o he mapped genes we e conduc ed
using KOBAS-i [43]. In all he analyses, a 5% alse disco e y a e (FDR)
was conside ed o mul iple es ing co ec ion.
SNP-based he i abili y
The p opo ion o pheno ypic a iance explained by SNPs was es ima ed
using wo di e en me hods. Fi s , SNP-based he i abili y was es ima ed
wi h he Linkage Disequilib ium Sco e Reg ession (LDSC) me hod [44]. To
educe he s anda d e o gi en ou ela i ely small sample size, he
in e cep was cons ained o 1 a e es ing ha i was no significan ly
highe han 1 [45]. The LDSC in e cep has been widely employed o
dis inguish be ween infla ion due o con ounding ac o s (such as
popula ion s a ifica ion and c yp ic ela edness) and infla ion due o
polygenici y. De ia ion o he in e cep om 1 is indica i e o esidual
con ounding, hus obse ing an in e cep no significan ly highe han 1
can be in e p e ed as he e being minimal con ounding bias [44]. Second,
we also es ima ed he i abili y using indi idual-le el geno ypes wi h he
Genome-based Res ic ed Maximum Likelihood (GREML) app oach imple-
men ed in he Genome-wide Complex T ai Analysis (GCTA) ool [46],
adjus ing o sex, he da ase and he op 10 PCs.
C oss- ai polygenic isk sco e
PRS analyses we e conduc ed using he PRS-cs so wa e [47] be ween en
men al pheno ypes and AAO. Specifically, we ob ained summa y s a is ics
da a on se en psychia ic diso de s downloaded om he Psychia ic
Genomics Conso ium (h ps://www.med.unc.edu/pgc/)[48], including
SCZ, BD, MDD, a en ion-defici /hype ac i i y diso de (ADHD), au ism
spec um diso de (ASD), obsessi e-compulsi e diso de (OCD), and
cannabis use diso de (CUD). In addi ion, we ob ained GWAS da a on
h ee condi ions p e iously associa ed wi h AAO: neu o icism [49],
educa ional a ainmen (EA) [11], and childhood mal ea men [50] (Table
S1). The downloaded summa y s a is ics we e fil e ed o emo e
ambiguous, mul iallelic and duplica ed SNPs, and SNPs wi h an impu a ion
sco e <0.9 (i he in o ma ion was p o ided). The summa y s a is ics o he
men al pheno ypes we e used as base da a, and he indi idual-le el
geno ype da ase was used as a ge da a. We calcula ed he PRS o each
men al pheno ype using he “au o”mode ( he sh inkage pa ame e phi
was de e mined om he da a wi h a Bayesian app oach). Then, he sco es
ob ained o each indi idual in ou da ase we e eg essed ou agains sex,
age, and ba ch o ob ain new adjus ed-PRS sco es. Linea eg essions we e
pe o med wi h he no malized AAO alues as ou come and he adjus ed-
PRSs as independen a iables o e alua e he associa ion o each PRS wi h
AAO. Finally, p- alues we e co ec ed o mul iple co ec ion using FDR.
Copy numbe a ia ion analysis
CNV analyses we e conduc ed using signal in ensi y da a om hose
indi iduals in he ou coho s om whom we ob ained he in ensi y files
(N=4630). The aw CNVs we e ob ained using PennCNV [51]. B iefly,
quali y con ol o CNV calls was based on a sample-le el c i e ion ha
examined he ela ionship be ween he s anda d de ia ion o he
loga i hm R Ra io (LRR_SD) and he numbe o CNV calls (NumCNV). A
he end o he p ocess, adjacen calls we e me ged oge he in o one
single call. Th esholds we e ca e ully chosen o include as many subjec s as
possible bu educe alse posi i es. Thus, subjec s wi h LRR_SD > 0.35,
BAF > 0.01, WF > 0.05, and NumCNV > 150 we e no included. Wi h all he
iden ified CNVs we pe o med linea eg ession analyses o es whe he
he AAO was associa ed wi h ei he he numbe o he o al leng h o
CNVs, dele ions o duplica ions. Sex, ba ch and he 10 fi s PCs we e used
as co a iables. In pa allel, 12 CNVs p e iously desc ibed as significan ly
associa ed wi h SCZ (SCZ-CNV) we e ob ained om he li e a u e [52]. The
BEDTools in e sec [53] was used o look o o e laps be ween he
desc ibed SCZ-CNVs and ou da a. B iefly, we selec ed as SCZ-CNV ca ie s
only hose indi iduals o whom a leas 90% o he SCZ-CNV o e lapped
wi h a de ec ed CNV (- 0.9) and/o wi h a leas 80% o ecip ocal o e lap
(- - 0.8). To de e mine whe he he p esence o SCZ-CNVs could be
influencing AAO, Wilcoxon and K uskal–Wallis es s we e pe o med in
R4.1.2 [54], using non-no malized AAO alues.
RESULTS
Age a onse
Mean AAO was 23.35 (SD =8.26). Mean AAO a ied ac oss
da ase s, anging om 21.11 o 26.26 (Table 1). In he whole
sample 1525 subjec s (32.4%) we e emales (mean AAO =25.09),
and 3182 subjec s (67.6%) we e males (mean AAO =22.52).
Significan di e ences we e de ec ed be ween he AAO o
E. Sada-Fuen e e al.
3
T ansla ional Psychia y (2023) 13:201
CIBERSAM and GAIN (P- alue < 2.2e−16), CIBERSAM and nonGAIN
(P- alue < 2.2e−16), PsyCou se and GAIN (P- alue =2.2e−16), and
PsyCou se and nonGAIN (P- alue =4.44e−16). Howe e , di e -
ences we e no de ec ed be ween CIBERSAM and PsyCou se (P-
alue =0.11), and nei he be ween GAIN and nonGAIN (P-
alue =0.33, Fig. S1). Since he dis ibu ion o AAO was igh -
skewed (Fig. S1), i was no malized using a ank-based in e se-
no mal ans o ma ion and used in all subsequen analyses.
GWAS and unc ional analyses
A o al o 4740 subjec s o Eu opean ances y and 6,540,522 SNPs
we e included in he GWAS me a-analysis. Al hough none o he
analyzed SNPs eached he genome-wide significan h eshold (P-
alue < 5e−08, Fig. 1), 25 lead SNPs we e iden ified, co espond-
ing o 22 genomic isk loci ha we e mapped o 183 genes (Table
2). Using he Linkage Disequilib ium Sco e Reg ession (LDSC)
me hod [44], an in e cep o 1.03 (SE =0.0083) was ob ained.
Mapped genes we e en iched in ca ego ies such as anspo o
small molecules (FDR-adj P- alue =1.7e−02), esicle-media ed
anspo (FDR-adj P- alue =1.8e−02), me abolism (FDR-adj P-
alue =2.55e−02), MHC class II an igen p esen a ion (FDR-adj P-
alue =2.6e−02), and Aspa agine N-linked glycosyla ion (FDR-adj
P- alue =3.3e−02), among o he s (Table S2).
SNP-based he i abili y
The SNP-based he i abili y (h
2SNP
) was es ima ed using wo
me hods ha showed consis en esul s wi h mode a e and
significan SNP-based he i abili y. Fi s , wi h LDSC, we ob ained
h
2SNP
=0.21 (SE =0.07). SNP-based he i abili y was also es ima ed
using indi idual-le el geno ypes wi h GCTA-GREML, adjus ing o
sex, da ase and he op 10 PCs, esul ing in an es ima e o
h
2SNP
=0.17 (SE =0.06; P- alue =3.33e−03). These esul s we e
consis en wi h he he i abili y es ima es ob ained om he GWAS
me ged app oach (h
2SNP
=0.13, Supplemen a y No e).
C oss- ai polygenic isk sco e
Ten men al pheno ypes we e examined h ough a c oss- ai PRS
analysis. Only he PRSs calcula ed based on ADHD, SCZ and
childhood mal ea men sums a s we e associa ed wi h AAO in
ou da ase (FDR < 0.05). All adjus ed PRS be a coe ficien s we e
nega i e, co esponding o he highe bu den o disease/
condi ion isk a ian s being associa ed wi h ea lie onse o SCZ
(Fig. 2and Table S3). The a iance explained by he adjus ed-PRS
o hese h ee pheno ypes was low bu significan (1.2e−03; 1.3e
−03 and 1.1e−03, espec i ely o ADHD, SCZ, and childhood
mal ea men ). Among he o he men al pheno ypes, only BD-PRS
was nominally associa ed wi h AAO (P=0.02).
CNV analysis
A e quali y con ol, a o al o 3965 indi iduals emained o CNV
downs eam analyses and a o al o 107,668 dele ions (mean
numbe =27.18 ± 17.12; mean leng h =1434 kb ± 1617 kb) and
59,336 duplica ions (mean numbe =15.05 ± 10.09; mean leng h
=2094 kb ± 6817 kb) we e iden ified. 3962 indi iduals (99.92%)
ca ied a leas one dele ion and 3942 (99.41%) indi iduals ca ied
a leas one duplica ion. AAO was significan ly associa ed wi h he
o al numbe o CNVs (be a =−0.015; P- alue =0.025), and also
wi h he numbe o dele ions (be a =−0.02; P- alue =0.03) and
he leng h o he dele ions (be a =−1.8e−7; P- alue =0.03, Table
3). Ou o 3965, 117 indi iduals (3%) ca ied SCZ-CNVs p e iously
associa ed wi h SCZ, and 4 indi iduals ca ied wo SCZ-CNVs.
Among he SCZ-CNV-ca ie s, we de ec ed 36 (29.8%) ca ying
15q11.2del, 20 (16.53%) 22q11.2del, 13 (10.74%) 16p12.1del, 12
(9.92%) 16p11.2dup, 11 (9.09%) 16p13.11dup, 10 (8.26%)
15q13.3del, 6 (4.96%) 3q29del, 4 (3.31%) ca ying 1q21.1dup, 4
(3.31%) 15q11-q13dup, 3 (2.48%) 1q21.1del, and 2 (1.65%)
7q11.23dup indi iduals. The dele ion co esponding o 2p16.3
was no p esen in ou sample. In ou da ase , he p esence o
SCZ-CNVs was no associa ed wi h an ea lie AAO (Wilcoxon es ;
P- alue =0.73, Fig. S2A). In addi ion, no di e ences in he AAO
we e ound ac oss SCZ-CNVs (K uskal–Wallis es ; P- alue =0.49,
Fig. S2B).
DISCUSSION
This s udy explo ed he gene ic a chi ec u e o AAO o SCZ. To his
end, we pe o med a case-only GWAS o Eu opean ances y in he
la ges sample collec ed o da e. Al hough no genome-wide
significan signals we e de ec ed, we success ully es ima ed o
he fi s ime he SNP-based he i abili y o AAO using wo di e en
app oaches ha showed consis en esul s o mode a e he i -
abili y, anging om 17 o 21%. The ac ha we did no iden i y
Fig. 1 Manha an plo (le ) and Q-Q plo (uppe igh ) o he genome-wide me a-analysis on AAO in SCZ. Red dashed line ep esen s he
h eshold o genome-wide significan associa ions (P- alue < 5e−08).
E. Sada-Fuen e e al.
4
T ansla ional Psychia y (2023) 13:201
Table 2. Genomic isk loci o AAO iden ified in he genome-wide me a-analysis.
Genomic Locus Posi ion
a
Lead SNP Lead SNP posi ion P- alue nSNPs
b
Genes
1 1:181854925 s185188889 1:181854925 6.735e−06 1 –
2 2:81911239:82893571 s13383639 82802994 7.8040E−06 19 REG3G, REG1B, REG3A, CTNNA2, LRRTM1, SUCLG1
3 3:173494302:173563936 s7652242 173553275 4.57E−06 59 NLGN1, NAALADL2
4 4:30196116:30196116 s111289733 30196116 1.66E−07 1 RP11-180C1.1
5 4:38573143:38604331 s4833071 38582859 4.79E−06 12 RELL1, TBC1D1, PTTG2, AC021860.1, KLF3, TLR10, RFC1, UGDH
6 4:40443966:40443966 s10755175 40443966 4.49E−06 1 RHOH, RBM47, NSUN7
7 4:167064039:167096817 s10018884 167092502 2.36E−06 65 MARCH1, MSMO1, SPOCK3, ANXA10, DDX60, PALLD
8 5:108260989:108520863 s78438786 108260989 6.72E−07 6 FER, PJA2, MAN2A1, TMEM232, SLC25A46
9 5:120718453:120901978 s2195409 120892594 6.73E−06 34 DMXL1, HSD17B4, FAM170A, FTMT, SRFBP1, LOX, SNCAIP
10 5:133457000:133774168 s4431386 133557876 2.54E−06 65 SLC22A5, C5o 56, IL4, KIF3A, CCNI2, GDF9, UQCRQ, LEAP2, AFF4, ZCCHC10,
HSPA4, C5o 15, VDAC1, TCF7, SKP1, CTD-2410N18.5, PPP2CA, CDKL3, UBE2B,
CDKN2AIPNL, JADE2, SAR1B, SEC24A, DDX46, C5o 24, TXNDC15, PCBD2, PITX1
11 7:48751259:48840965 s139864446 48831950 9.07E−06 7 AC004899.1, VWC2, C7o 72, IKZF1, DDC
12 7:106206611:106399401 s111513327 106207969 1.75E−06 7 EFCAB10, ATXN7L1, SYPL1, NAMPT, CCDC71L, PIK3CG, PRKAR2B, HBP1, COG5,
DUS4L, BCAP29, SLC26A4, CBLL1, SLC26A3
13 8:20289797:20370404 s12550821 20315601 4.85E−06 32 CSGALNACT1, LPL, SLC18A1, ATP6V1B2, LZTS1, GFRA2, DOK2, XPO7, LGI3, SFTPC,
BMP1
14 8:121006828:121262332 s10808509 121227216 5.61E−06 77 TAF2, DSCC1, DEPTOR, COL14A1, MRPL13, MTBP, SNTB1
15 11:2192798:2192798 s10840489 2192798 2.04E−06 1 KRTAP5-4, KRTAP5-5, KRTAP5-6, IFITM10, RP11-295K3.1, CTSD, SYT8, TNNI2, LSP1,
C11o 89, TNNT3, TH, C11o 21
16 11:44644743:44870058 s11038082 44644743 2.51E−06 28 C11o 96, EXT2, ALX4, CD82, TSPAN18, TP53I11, PRDM11, SYT13, SLC35C1, CRY2
17 12:32500207:32537185 s144642024 32500207 1.58E−06 2 OVCH1, FAM60A, AC024940.1, DENND5B, METTL20, AMN1, KIAA1551, FGD4,
DNM1L, YARS2, ALG10
18 13:61356260:61390956 s9570366 61377708 1.69E−06 19 PCDH20
19 14:20479798:20546983 s72667672 20492904 6.82E−06 82 OR4N2, OR4K2, OR4K5, OR4K1, OR4K14, OR4K13, OR4L1, OR4K17, OR4N5, TTC5,
RNASE9, RNASE4, ANG, AL163636.6
20 16:71388416:71388416 s142248381 71388416 2.45E−07 1 COG4, SF3B3, MTSS1L, VAC14, HYDIN, CMTR2, ZNF23, ZNF19, TAT, HP, HPR
21 16:84318421:84322649 s11645140 84322546 5.17E−07 4 NECAB2, MBTPS1, HSDL1, DNAAF1, KCNG4, WFDC1, ATP2C2
22 17:51482920:52487345 s146709267 52065109 1.48E−06 11 MBTD1, UTP18, CA10, AC102948.2, C17o 112, KIF2B, TOM1L1, COX11, STXBP4,
HLF, MMD, ANKFN1, NOG
a
Posi ions a e based on Human Genome e sion 19 (hg19), build 37.
b
Numbe o SNPs in he genomic locus (
2
≥0.6 wi h any o he independen significan SNPs).
E. Sada-Fuen e e al.
5
T ansla ional Psychia y (2023) 13:201
any specific gene ic a ian s associa ed wi h AAO sugges s ha
he he i abili y o he ai is likely o be complex, in ol ing
mul iple gene ic and en i onmen al ac o s. We also p o ided
e idence o nega i e gene ic associa ions o c oss- ai PRS
de i ed om ADHD, SCZ, and childhood mal ea men wi h
AAO. Finally, we de e mined ha he bu den o dele ions was
associa ed wi h AAO in ou da ase .
In ou s udy, he s onges associa ion signal was ound in a
genomic locus a ch omosome 4 (lead SNP s111289733 P-
alue =1.66e−07), which ha bo ed he long non-coding RNA
RP11-180C1.1. The mapped genes belonging o he sugges i e
associa ions we e en iched in anspo o small molecules,
esicle-media ed anspo , Aspa agine N-linked glycosyla ion,
and MHC class II an igen p esen a ion, among o he s. Bo h he
anspo o molecules and, specifically, he esicula anspo
mechanism migh pa icipa e in he pa hogenesis o SCZ by
igge ing dys unc ional neu oexocy osis [55,56]. In ac , p e ious
s udies epo ed abno mal educ ions in synap ic esicle p o eins
being associa ed wi h SCZ [57–59]. Also, MHC has been s ongly
associa ed wi h he isk o SCZ [60]. In addi ion, glycosyla ion
in ol es p ocesses c i ical o no mal b ain de elopmen and has
been sugges ed o con ibu e o he abno mal neu onal signaling
and connec i i y obse ed in SCZ [61]. Al oge he , hese
ca ego ies a e p omising candida es o u he s udies on
pa hways associa ed wi h AAO.
O e he yea s, many s udies ha e e alua ed he ole o
gene ics in AAO o SCZ [62,63], es ima ing a he i abili y o AAO
anging om app oxima ely 20 o 58% [7]. In ou s udy, based on
wo di e en me hodologies we es ima ed he SNP-based
he i abili y o AAO o be be ween 17 and 21%. In ac , ou
es ima es show consis en esul s sugges ing a mode a e bu
significan con ibu ion o common a ian s o AAO. In e es ingly,
SNP-based he i abili y is sligh ly highe han ha o AAO in BD
and MDD, which has ecen ly been es ima ed a 5 and 6%
espec i ely, using la ge sample sizes [27]. Fu he s udies wi h
la ge sample sizes a e needed o ob ain mo e accu a e es ima es
in SCZ AAO.
C oss- ai PRSs cons uc ed wi h ADHD, SCZ, and childhood
mal ea men we e significan ly associa ed wi h AAO, showing
ha a highe isk o de eloping hese condi ions is associa ed wi h
an ea lie AAO in SCZ. Howe e , hey explained a e y small
ac ion o AAO a ia ion [64]. P e ious s udies epo ed ha
ADHD was among he commones como bidi ies in child en and
adolescen s wi h SCZ [65], and i has been a gued ha he gene ic
a chi ec u e o ADHD has a la ge link wi h SCZ [66]. In his line, ou
esul s sugges ha an ea lie AAO may be ela ed o a mo e
se e e neu ode elopmen al impai men . AAO has been sugges ed
as a po en ial endopheno ype o SCZ, eflec ing he unde lying
gene ic a chi ec u e o he diso de . Ou finding ha he PRS o
SCZ can p edic AAO suppo s his no ion and sugges s ha AAO
could be a use ul p edic o o disease se e i y. I has also been
sugges ed ha indi iduals wi h highe gene ic loadings o SCZ
a e a a highe isk o ea ly onse [67], and simila ly o MDD [28].
Mo eo e , pa ien s wi h his o ies o being abused as child en
show an ea lie onse o symp oms [68]. He e, we also epo a
nega i e associa ion be ween AAO and he PRS o bo h SCZ and
childhood mal ea men . Howe e , i is s ill unknown how he
gene ic a chi ec u es o hese ai s a e linked.
A ecen s udy epo ed ha he p e alence o ecu en CNVs
was highe in ea ly onse psychosis han in he gene al
popula ion, as well as CNV pa hogenici y [67]. In addi ion, some
o hese CNVs cause ea lie -onse diso de s such as de elopmen al
delay o ASD, bu no SCZ [20]. He e, we obse ed ha he bu den
o CNVs, especially dele ions (ei he he leng h o he o al
numbe ), bu no duplica ions, we e associa ed wi h an ea lie
onse o SCZ, sugges ing a combined ole o common a ia ion
and CNVs in de e mining AAO. Howe e , in ou s udy, nei he he
p esence o pa hogenic SCZ-CNV, no any specific SCZ-CNVs we e
associa ed wi h AAO. Ne e heless, we de ec ed a 1.9% p e-
alence o he pa hogenic SCZ-CNVs, which is close o he
p e iously epo ed p e alence o 2.6% [20].
Ou s udy has some s eng hs and limi a ions ha dese e
discussion. Despi e he lack o genome-wide significan associa-
ions a he SNP-le el, which could be expec ed gi en he
ela i ely small sample size o ou s udy, we we e able o es ima e
he SNP-based he i abili y o AAO and iden i y ele an associa-
ions wi h he PRS o h ee men al pheno ypes o he fi s ime.
This finding ep esen s an impo an s ep owa d elucida ing he
gene ic a chi ec u e o AAO. I is wo h no ing ha s udying AAO
is bo h echnically and concep ually challenging, as he e is o en
conside able a iabili y in how AAO is defined and measu ed
be ween s udies. Al hough hese measu es ha e been shown o
be co ela ed and o occu wi hin a ela i ely sho ime ame
[6,69], i is possible ha ou esul s may be con ounded by
di e en defini ions o AAO be ween da ase s. In ou s udy, we
acknowledge ha AAO di e ed be ween he Eu ope and US
coho s in ou s udy and he e o e pe o med a GWAS me a-
analysis o minimize any po en ial bias due o pheno ype
he e ogenei y. Howe e , he ac ha AAO did no di e wi hin
he Eu opean (CIBERSAM and PsyCou se) and he US (GAIN and
nonGAIN) coho s sugges s ha he pheno ypes we e ob ained in
a homogeneous manne . In he u u e, mo e eliable es ima es o
AAO may lead o he iden ifica ion o significan signals and an
inc ease in he he i abili y explained by SNPs. Also, he ank-based
ans o ma ion applied o he AAO alues may ha e a ec ed ou
GWAS and subsequen analyses; howe e , i is conside ed as one
o he bes app oaches o use. Some s udies ha e epo ed ha
o small sample sizes o gene ic e ec s, he e is an imp o emen
in sensi i i y o ank-based ans o ma ions ha ou weighs a
sligh inc ease in he alse-posi i e a e [70]. In addi ion, a ecen
s udy has demons a ed ha hese ans o ma ion es s
Fig. 2 Resul s o he associa ions o adjus ed-PRSs wi h AAO. Be a
es ima es (x-axis) o he adjus ed-PRS o he linea model a e
shown. Non-significan alues a e colo ed in g ey (FDR > 0.05) and
significan esul s (FDR < 0.05) in ed. Adjus ed R
2
alues o he linea
models a e shown o each pheno ype on he igh side o he
panel. SCZ Schizoph enia, OCD obsessi e-compulsi e diso de , Neu
Neu o icism, MDD majo dep essi e diso de , EA educa ional
a ainmen , CUD Cannabis use diso de , CM childhood mal ea -
men , BD bipola diso de , ASD Au ism spec um diso de , ADHD
a en ion-defici /hype ac i i y diso de .
E. Sada-Fuen e e al.
6
T ansla ional Psychia y (2023) 13:201
Table 3. Resul s om linea models o leng h and numbe o CNVs, dele ions and duplica ions.
Numbe o CNVs Numbe o dele ions Numbe o duplica ions
Be a PBe a PBe a P
Numbe −0.015 0.025 −0.017 0.029 0.009 0.492
PC1 −15.50 0.054 −15.947 0.046 −15.786 0.048
PC2 −8.43 0.293 −9.760 0.221 −9.682 0.225
PC3 1.91 0.813 0.643 0.936 1.295 0.872
PC4 −3.43 0.673 −1.059 0.896 0.593 0.941
PC5 5.07 0.523 4.505 0.568 3.521 0.655
PC6 7.17 0.366 5.659 0.473 5.314 0.501
PC7 −3.21 0.683 −2.753 0.725 −3.220 0.681
PC8 9.09 0.249 9.584 0.221 9.817 0.211
PC9 1.45 0.853 1.507 0.846 1.499 0.847
PC10 −5.64 0.476 −6.727 0.392 −6.771 0.389
Sex 2.43 <2e−16 −2.406 <2e−16 2.382 <2e−16
Ba ch −1.24 <2e−16 −3.265 <2e-16 −3.550 <2e−16
adj-R
2
0.065 0.075 0.075
Model p al <2.2e
−16
<2.2e
−16
<2.2e
−16
Leng h o CNVs Leng h o dele ions Leng h o duplica ions
Be a PBe a PBe a P
Leng h -4.77E-09 0.798 0.000 0.029 0.000 0.809
PC1 −15.74 0.049 −0.160 0.045 −0.157 0.049
PC2 −9.63 0.228 −9.650 0.226 −9.693 0.225
PC3 1.07 0.894 1.480 0.854 1.225 0.879
PC4 0.60 0.941 −0.010 0.990 0.441 0.956
PC5 3.59 0.649 4.010 0.610 3.618 0.646
PC6 5.26 0.505 5.810 0.462 5.289 0.503
PC7 −3.16 0.686 −3.015 0.700 −3.181 0.684
PC8 9.88 0.208 10.050 0.200 9.688 0.217
PC9 1.51 0.846 1.760 0.821 1.482 0.849
PC10 −6.77 0.390 −6.745 0.391 −6.779 0.389
Sex 2.39 <2e−16 2.490 <2e−16 2.380 <2e−16
Ba ch −3.47 <2e−16 −3.250 <2e−16 −3.522 <2e−16
adj-R
2
0.074 0.075 0.074
Model p al <2.2e
−16
<2.2e
−16
<2.2e
−16
Bold alues iden i y s a is ical significance (p< 0.05).
E. Sada-Fuen e e al.
7
T ansla ional Psychia y (2023) 13:201
ou pe o m he s anda d un ans o med associa ion es , bo h in
e ms o powe and ype I e o a e con ol [71]. Mo eo e , we
we e no able o con ol o pu a i e di e ences be ween he
di e en ec ui men si es. All he da ase s included in he s udy
migh be mul icen e ; hus, he e ogenei y wi hin da ase s could
be conside able. Such pheno ypic he e ogenei y has been
epo ed o a ec gene ic analyses [27,72], which indica es ha
pheno ype ha moniza ion is as impo an as a la ge sample size
o imp o ing he powe o de ec significan associa ions and
a oiding a biased iew o gene ic a chi ec u es.
CONCLUSIONS
In conclusion, we epo on he la ges GWAS o AAO in SCZ o
da e, p o iding he fi s SNP-based he i abili y es ima e o AAO
in indi iduals o Eu opean ances y. Al hough no genome-wide
significan SNP was de ec ed, we p o ide e idence o a gene ic
backg ound o AAO and a nega i e associa ion wi h he PRS o
ADHD, SCZ and childhood mal ea men . In addi ion, we
demons a e ha he bu den o dele ions is associa ed wi h
he AAO o he disease. La ge sample sizes a e needed o ully
de e mine he gene ic a chi ec u e o AAO, which could help us
unde s and u he he pa hogenesis o SCZ and con ibu e o
he de elopmen o be e s a egies o he ea ly de ec ion o
SCZ. None heless, ou s udy p o ides an impo an s ep o wa d
in unde s anding he gene ic a chi ec u e o AAO.
DATA AVAILABILITY
Gi en he collabo a i e na u e o he CIBERSAM coho and he in ol emen o
mul iple ins i u ions and esea ch g oups, eques s o access o he geno ype da a
should be add essed o he co esponding au ho [mun ane[email p o ec ed]] o
ini ia e he da a access eques . Acces o Psycou se da ase can be eques ed ho ugh
a“Seconda y Analysis P oposal”a h p://www.psycou se.de/openscience-en.h ml.
REFERENCES
1. Delisi LE. The significance o age o onse o schizoph enia. Schizoph Bull.
1992;18:209–15.
2. Öngü D, Lin L, Cohen BM. Clinical cha ac e is ics influencing age a onse in
psycho ic diso de s. Comp Psychia y. 2009;50:13–9.
3. Kendle KS, MacLean CJ. Es ima ing amilial e ec s on age a onse and liabili y o
schizoph enia. I. Resul s o a la ge sample amily s udy. Gene Epidemiol.
1990;7:409–17.
4. Sham PC, Jones P, Russell A, Gil a y K, Bebbing on P, Lewis S, e al. Age a onse ,
sex, and amilial psychia ic mo bidi y in schizoph enia. Cambe well collabo a i e
psychosis s udy. B J Psychia y. 1994;165:466–73.
5. Rajji TK, Ismail Z, Mulsan BH. Age a onse and cogni ion in schizoph enia: me a-
analysis. B J Psychia y. 2009;195:286–93.
6. Immonen J, Jääskeläinen E, Ko pela H, Mie unen J. Age a onse and he ou -
comes o schizoph enia: a sys ema ic e iew and me a-analysis: age a onse and
he ou comes o schizoph enia. 2017. h ps://doi.o g/10.1111/eip.12412.
7. Muske CW,KuoSS,Rupe PE,AlmasyL,Gu RC,P asadK,e al.Whydoesage
o onse p edic clinical se e i y in schizoph enia? A mul iplex ex ended
pedig ee s udy. Am J Med Gene , Pa B: Neu opsychia ic Gene .
2020;183:403–11.
8. Aleman A, Kahn RS, Sel en JP. Sex di e ences in he isk o schizoph enia: e i-
dence om me a-analysis. A chi es Gene al Psychia y. 2003;60:565–71.
9. Leung M. Sex di e ences in schizoph enia, a e iew o he li e a u e. Ac a Psy-
chia Scand Suppl. 2000;401:3–38.
10. Neill E, Tan EJ, Lin Toh W, Sel end a A, Mo gan VA, Rossell SL, e al. Examining
which ac o s influence age o onse in males and emales wi h schizoph enia.
2020. h ps://doi.o g/10.1016/j.sch es.2020.08.011.
11. Ochoa S, Usall J, Cobo J, Labad X, Kulka ni J. Gende di e ences in schizoph enia
and fi s -episode psychosis: a comp ehensi e li e a u e e iew. Schizoph Res.
T ea men . 2012;2012:1–9.
12. Sulli an PF, Kendle KS, Neale MC. Schizoph enia as a complex ai : e idence
om a me a-analysis o win s udies. A chi es Gene al Psychia y.
2003;60:1187–92.
13. Lich ens ein P, Yip BH, Bjö k C, Pawi an Y, Cannon TD, Sulli an PF, e al. Common
gene ic influences o schizoph enia and bipola diso de : a popula ion-based
s udy o 2 million nuclea amilies. Lance . 2009;373:1–14.
14. Conso ium TSWG o he PG, Ripke S, Wal e s JT, O’Dono an MC. Mapping
genomic loci p io i ises genes and implica es synap ic biology in schizoph enia.
medRxi 2020; 2020.09.12.20192922.
15. Le inson DF, Duan J, Oh S, Wang K, Sande s AR, Shi J, e al. Copy numbe a ian s
in schizoph enia: Confi ma ion o fi e p e ious finding sand new e idence o
3q29 mic odele ions and VIPR2 duplica ions. Am J Psychia y. 2011;168:302–16.
16. Rees E, Wal e s JTR, Geo gie a L, Isles AR, Chambe KD, Richa ds AL, e al.
Analysis o copy numbe a ia ions a 15 schizoph enia-associa ed loci. B J
Psychia y. 2014;204:108–14.
17. Rees E, Kendall K, Pa diñas AF, Legge SE, Pockling on A, Esco -P ice V, e al.
Analysis o in ellec ual disabili y copy numbe a ian s o associa ion wi h
schizoph enia. JAMA Psychia y. 2016;73:963–9.
18. Ma shall CR, How igan DP, Me ico D, Thi u ahind apu am B, Wu W, G ee DS,
e al. Con ibu ion o copy numbe a ian s o schizoph enia om a genome-
wide s udy o 41,321 subjec s. Na Gene . 2016;49:27–35.
19. Hal o sen M, Huh R, Oskolko N, Wen J, Ne o ea S, Gius i-Rod iguez P, e al.
Inc eased bu den o ul a- a e s uc u al a ian s localizing o bounda ies o
opologically associa ed domains in schizoph enia. Na Commun. 2020;11:1–13.
20. Ki o G, Rees E, Wal e s JTR, Esco -P ice V, Geo gie a L, Richa ds AL, e al. The
pene ance o copy numbe a ia ions o schizoph enia and de elopmen al
delay. Biol Psychia y. 2014;75:378–85.
21. Ha e E, Glahn DC, Dasso i A, Ra en os H, Nicolini H, On i e os A, e al. He i abili y
o age o onse o psychosis in schizoph enia. Am J Med Gene , Pa B: Neu-
opsychia ic Gene . 2010;153:298–302.
22. Ripke S, O’Dushlaine C, Chambe K, Mo an JL, Kähle AK, Ak e in S, e al.
Genome-wide associa ion analysis iden ifies 13 new isk loci o schizoph enia.
Na Gene . 2013;45:1150–9.
23. Wo king G oup o he Psychia ic Genomics Conso ium S. Biological insigh s
om 108 schizoph enia-associa ed gene ic loci. 2014. h ps://doi.o g/10.1038/
na u e13595.
24. Wang K-S, Liu X, Zhang Q, A agam N, Pan Y. Genome-wide associa ion analysis o
age a onse in schizoph enia in a Eu opean-Ame ican sample. Am J Med Gene
Pa B: Neu opsychia ic Gene . 2011;156:671–80.
25. Be gen SE, O’Dushlaine CT, Lee PH, Fanous AH, Rude e DM, Ripke S, e al.
Gene ic modifie s and sub ypes in schizoph enia: In es iga ions o age a onse ,
se e i y, sex and amily his o y. Schizoph Res. 2014;154:48–53.
26. Wools on AL, Hsiao P-CC, Kuo P-HH, Wang S-HH, Lien Y-JJ, Liu C-MM, e al.
Gene ic loci associa ed wi h an ea lie age a onse in mul iplex schizoph enia. Sci
Rep. 2017;7:6486.
27. Kalman JL, Olde Loohuis LM, V eeke A, McQuillin A, S ahl EA, Rude e D, e al.
Cha ac e isa ion o age and pola i y a onse in bipola diso de . B J Psychia y.
2021;219:659–69.
28. Ha de A, Nguyen T-D, Pasman JA, Mosing MA, Hägg S, Lu Y. Gene ics o age-a -
onse in majo dep ession. T ansl Psychia y. 2022;12:1–7.
29. The GAIN Collabo a i e Resea ch G oup. New models o collabo a ion in
genome-wide associa ion s udies: he Gene ic Associa ion In o ma ion Ne wo k.
Na Gene . 2007;39:1045–51.
30. Salag e E, A ango C, A igas F, Ayuso-Ma eos JL, Be na do M, Cas o-Fo nieles J,
e al. CIBERSAM: Ten yea s o collabo a i e ansla ional esea ch in men al dis-
o de s. Re is a de Psiquia ía y Salud Men al (English Edi ion). 2019;12:1–8.
31. T ube skoy V, Pa diñas AF, Qi T, Panagio a opoulou G, Awas hi S, Bigdeli TB, e al.
Mapping genomic loci implica es genes and synap ic biology in schizoph enia.
Na u e. 2022;604:502–8.
32. Budde M, Ande son-Schmid H, Gade K, Reich-E kelenz D, Ado jan K, Kalman JL,
e al. A longi udinal app oach o biological psychia ic esea ch: The PsyCou se
s udy. Am J Med Gene , Pa B: Neu opsychia ic Gene . 2019;180:89–102.
33. Shi J, Le inson DF, Duan J, Sande s AR, Zheng Y, Pée I, e al. Common a ian s on
ch omosome 6p22.1 a e associa ed wi h schizoph enia. Na u e. 2009;460:753–7.
34. Pu cell S, Neale B, Todd-B own K, Thomas L, Fe ei a MAR, Bende D, e al. PLINK:
a ool se o whole-genome associa ion and popula ion-based linkage analyses.
Am J Hum Gene . 2007;81:559–75.
35. Ma ees AT, de Klui e H, S inge S, Vo span F, Cu is E, Ma ie-Clai e C, e al. A
u o ial on conduc ing genome-wide associa ion s udies: Quali y con ol and
s a is ical analysis. In J Me hods Psychia Res. 2018;27:e1608.
36. Pa e son N, P ice AL, Reich D. Popula ion s uc u e and eigenanalysis. PLoS
Gene . 2006;2:e190.
37. 1000 Genomes P ojec Conso ium, Au on A, B ooks LD, Du bin RM, Ga ison EP,
Kang HM, e al. A global e e ence o human gene ic a ia ion. Na u e.
2015;526:68–74.
38. NHLBI T ans-Omics o P ecision Medicine (TOPMed) Conso ium, Taliun D, Ha is
DN, Kessle MD, Ca lson J, Szpiech ZA, e al. Sequencing o 53,831 di e se gen-
omes om he NHLBI TOPMed P og am. Na u e. 2021;590:290–9.
39. Na a o Gonzalez J, Zweig AS, Spei ML, Schmel e D, Rosenbloom KR, Raney BJ,
e al. The UCSC Genome B owse da abase: 2021 upda e. Nucleic Acids Res.
2021;49:D1046–57.
E. Sada-Fuen e e al.
8
T ansla ional Psychia y (2023) 13:201
40. Wille CJ, Li Y, Abecasis GR. METAL: as and e ficien me a-analysis o genome-
wide associa ion scans. Bioin o ma ics. 2010;26:2190–1.
41. Wa anabe K, Taskesen E, Van Bocho en A, Pos huma D. Func ional mapping and
anno a ion o gene ic associa ions wi h FUMA. Na Commun. 2017;8:1–11.
42. Wang K, Li M, Hakona son H. ANNOVAR: Func ional anno a ion o gene ic a -
ian s om high- h oughpu sequencing da a. Nucleic Acids Res. 2010;38:1–7.
43. Bu D, Luo H, Huo P, Wang Z, Zhang S, He Z, e al. KOBAS-i: in elligen p io -
i iza ion and explo a o y isualiza ion o biological unc ions o gene en ichmen
analysis. Nucleic Acids Res. 2021;49. h ps://doi.o g/10.1093/na /gkab447.
44. Bulik-Sulli an B, Loh PR, Finucane HK, Ripke S, Yang J, Pa e son N, e al. LD sco e
eg ession dis inguishes con ounding om polygenici y in genome-wide asso-
cia ion s udies. Na Gene . 2015;47:291–5.
45. Nie e gel CM. In e na ional me a-analysis o PTSD genome-wide associa ion
s udies iden ifies sex-and ances y-specific gene ic isk loci. h ps://doi.o g/
10.1038/s41467-019-12576-w.
46. Yang J, Lee SH, Godda d ME, Vissche PM. GCTA: A ool o genome-wide com-
plex ai analysis. Am J Hum Gene . 2011;88:76–82.
47. Ge T, Chen C-Y, Ni Y, Feng Y-CA, Smolle JW. Polygenic p edic ion ia Bayesian
eg ession and con inuous sh inkage p io s. Na Commun. 2019;10:1776.
48. Sulli an PF, Ag awal A, Bulik CM, And eassen OA, Bø glum AD, B een G, e al.
Psychia ic genomics: an upda e and an agenda. h ps://doi.o g/10.1176/
appi.ajp.2017.17030283.
49. Goodwin RD, Fe gusson DM, Ho wood LJ. Neu o icism in adolescence and psy-
cho ic symp oms in adul hood. Psychol Med. 2003;33:1089–97.
50. Va ese F, Smee s F, D ukke M, Lie e se R, La as e T, Viech baue W, e al.
Childhood ad e si ies inc ease he isk o psychosis: A me a-analysis o pa ien -
con ol, p ospec i e-and c oss-sec ional coho s udies. Schizoph Bull.
2012;38:661–71.
51. Wang K, Li M, Hadley D, Liu R, Glessne J, G an SFA, e al. PennCNV: an in eg a ed
hidden Ma ko model designed o high- esolu ion copy numbe a ia ion de ec-
ion in whole-genome SNP geno yping da a. h ps://doi.o g/10.1101/g .6861907.
52. Wa land A, Kendall KM, Rees E, Ki o G, Case as X. Schizoph enia-associa ed
genomic copy numbe a ian s and subco ical b ain olumes in he UK Biobank.
Mol Psychia y. 2020;25:854–62.
53. Quinlan AR, Hall IM. BEDTools: a flexible sui e o u ili ies o compa ing genomic
ea u es. Bioin o ma ics. 2010;26:841–2.
54. Team RC. R: A language and en i onmen o s a is ical compu ing. 2021.
55. Egbujo CN, Sinclai D, Hahn C-G. Dys egula ions o synap ic esicle a ficking in
schizoph enia. Cu Psychia y Rep. 2016;18:77.
56. Schube KO, Föcking M, P ehn JHM, Co e DR. Hypo hesis e iew: a e cla h in-
media ed endocy osis and cla h in-dependen memb ane and p o ein a ficking
co e pa hophysiological p ocesses in schizoph enia and bipola diso de ? Mol
Psychia y. 2012;17:669–81.
57. Naka o M, Shi anaga N, Tomioka M, Wa anabe H, Ku isu J, Kengaku M, e al.
ABCA13 dys unc ion associa ed wi h psychia ic diso de s causes impai ed cho-
les e ol a ficking. J Biol Chem. 2021;296:100166.
58. Osimo EF, Beck K, Reis Ma ques T, Howes OD. Synap ic loss in schizoph enia: a
me a-analysis and sys ema ic e iew o synap ic p o ein and mRNA measu es.
Mol Psychia y. 2019;24:549–61.
59. Onwo di EC, Hal EF, Whi ehu s T, Mansu A, Co el M-C, Wells L, e al. Synap ic
densi y ma ke SV2A is educed in schizoph enia pa ien s and una ec ed by
an ipsycho ics in a s. Na Commun. 2020;11:246.
60. Mokh a i R, Lachman HM. The Majo His ocompa ibili y Complex (MHC) in
Schizoph enia: A Re iew. J Clin Cell Immunol. 2016;07. h ps://doi.o g/10.4172/
2155-9899.1000479.
61. Williams SE, Meale RG, Scolnick EM, Smolle JW, Cummings RD. Abe an gly-
cosyla ion in schizoph enia: a e iew o 25 yea s o pos -mo em b ain s udies.
Mol Psychia y. 2020;25:3198–207.
62. Va es M, Sae e P, Deng H, Cai G, Liu X, Hansen T, e al. Associa ion be ween
me hylene e ahyd o ola e educ ase (MTHFR) C677T polymo phism and age o
onse in schizoph enia. Am J Med Gene . 2010;153B:610–8.
63. Guo S, Liu J, Li W, Yang Y, L L, Xiao X, e al. Genome wide associa ion s udy
iden ifies ou loci o ea ly onse schizoph enia. T ansl Psychia y. 2021;11:248.
64. An ila V, Bulik-Sulli an B, Finucane HK, Wal e s RK, B as J, Duncan L, e al.
Analysis o sha ed he i abili y in common diso de s o he b ain. Science.
2018;360. h ps://doi.o g/10.1126/science.aap8757.
65. Ross RG, Heinlein S, T egellas H. High a es o como bidi y a e ound in
childhood-onse schizoph enia. Schizoph Res. 2006;88:90–5.
66. Hamshe e ML, S e giakouli E, Langley K, Ma in J, Holmans P, Ken L, e al. Sha ed
polygenic con ibu ion be ween childhood a en ion-defici hype ac i i y dis-
o de and adul schizoph enia. B J Psychia y. 2013;203:107–11.
67. Bea den CC, Glahn D, B owns ein C, Doua d E, Mollon J, Cab al K, e al. P e alence
o a e o dele e ious copy numbe a ian s simila in ea ly onse psychosis and
au ism spec um diso de s: implica ions o clinical p ac ice. Biol Psychia y.
2022;91:S56–7.
68. Kau man J, To bey S. Child mal ea men and psychosis. Neu obiol Dis.
2019;131:104378.
69. Jones PB. Adul men al heal h diso de s and hei age a onse . B J Psychia y.
2013;202:s5–s10.
70. Goh L, Yap VB. E ec s o no maliza ion on quan i a i e ai s in associa ion es .
BMC Bioin o ma ics. 2009;10. h ps://doi.o g/10.1186/1471-2105-10-415.
71. McCaw ZR, Lane JM, Saxena R, Redline S, Lin X. Ope a ing cha ac e is ics o he
ank-based in e se no mal ans o ma ion o quan i a i e ai analysis in
genome-wide associa ion s udies. Biome ics. 2020;76:1262–72.
72. Cai N, Re ez JA, Adams MJ, Andlaue TFM, B een G, By ne EM, e al. Minimal
pheno yping yields genome-wide associa ion signals o low specifici y o majo
dep ession. Na Gene . 2020;52:437–47.
ACKNOWLEDGEMENTS
This wo k was suppo ed by Ins i u o de Salud Ca los III h ough he p ojec s PI18/
00514 and PI21/00612 and co- unded by he Eu opean Union, by he Ca alan Agency
o Resea ch and Uni e si ies (AGAUR, 2017SGR-00444 and 2021SGR01065). The
PsyCou se s udy was suppo ed by DFG (SCHU 1603/4-1, 5-1, 7-1, FA241/16-1).
AUTHOR CONTRIBUTIONS
ES-F ca ied ou he analyses wi h he help o SA. GM concei ed and designed he
s udy. GM and EV, wi h he help o SP, JLK and TGS, pa icipa ed in he scien ific
discussion and supe ision o he p ojec . The o he au ho s con ibu ed o da a
collec ion. All au ho s c i ically e ised and app o ed he final e sion o he
manusc ip .
COMPETING INTERESTS
The au ho s decla e no compe ing in e es s.
ADDITIONAL INFORMATION
Supplemen a y in o ma ion The online e sion con ains supplemen a y ma e ial
a ailable a h ps://doi.o g/10.1038/s41398-023-02508-0.
Co espondence and eques s o ma e ials should be add essed o Elisabe Vilella o
Ge a d Mun ané.
Rep in s and pe mission in o ma ion is a ailable a h p://www.na u e.com/
ep in s
Publishe ’s no e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims
in published maps and ins i u ional a filia ions.
Open Access This a icle is licensed unde a C ea i e Commons
A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing,
adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e
app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e
Commons license, and indica e i changes we e made. The images o o he hi d pa y
ma e ial in his a icle a e included in he a icle’s C ea i e Commons license, unless
indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he
a icle’s C ea i e Commons license and you in ended use is no pe mi ed by s a u o y
egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly
om he copy igh holde . To iew a copy o his license, isi h p://
c ea i ecommons.o g/licenses/by/4.0/.
© The Au ho (s) 2023, co ec ed publica ion 2023
E. Sada-Fuen e e al.
9
T ansla ional Psychia y (2023) 13:201
