Regulatory variant of the TPH2 gene and early life stress are associated with heightened attention to social signals of fear in infants

Regula o y a ian o he TPH2 gene and ea ly li e s ess a e associa ed wi h heigh ened a en ion o social signals o ea in in an s Linda Fo ssman, 1 Mikko J. Pel ola, 2,3,4 San e i Y iaho, 1 Kaija Puu a, 5 Nina Mononen, 6 Te ho Leh im€ aki, 6 and Jukka M. Lepp€ anen 1 1 School o Medicine, Uni e si y o Tampe e, Tampe e, Finland; 2 School o Social Sciences and Humani ies, Uni e si y o Tampe e, Tampe e, Finland; 3 Cen e o Child and Family S udies, Leiden Uni e si y, Leiden, Ne he lands; 4 Leiden Ins i u e o B ain and Cogni ion, Leiden Uni e si y, Leiden, Ne he lands; 5 Depa men o Child Psychia y, Tampe e Uni e si y Hospi al, Tampe e, Finland; 6 Depa men o Clinical Chemis y, Fimlab Labo a o ies and School o Medicine, Uni e si y o Tampe e and Tampe e Uni e si y Hospi al, Tampe e, Finland Backg ound: C oss-species e idence sugges s ha gene ic and expe ien ial ac o s ac ea ly in de elopmen o es ablish indi idual emo ional ai s, bu li le is known abou he mechanisms ha eme ge du ing his pe iod o media e long- e m ou comes. He e, we es ed he hypo hesis ha known gene ic and en i onmen al isk condi ions may heigh en in an s’ na u al endency o a end o h ea -ale ing s imuli, esul ing in a cogni i e bias ha may con ibu e o emo ional ulne abili y. Me hods: Da a om wo samples o 5–7-mon h-old in an s (N=139) we e used o examine whe he es ablished candida e a ia ions in he se o onin-sys em genes, i.e., TPH2 SNP s4570625 (-703 G/T) and HTR1A SNP s6295 (-1019 G/C), and ea ly ea ing condi ion (ma e nal s ess and dep essi e symp oms) a e associa ed wi h al e a ions in in an s’ a en ion o acial exp essions. In an s we e es ed wi h a pa adigm ha assesses he abili y o disengage a en ion om a cen ally p esen ed s imulus (a non ace con ol s imulus o a neu al, happy, o ea ul acial exp ession) owa d he loca ion o a new s imulus in he isual pe iphe y (a geome ic shape). Resul s: TPH2 -703 T-ca ie geno ype (i.e., TT homozygo es and he e ozygo es), p esence o ma e nal s ess and dep essi e symp oms, and a combina ion o he T-ca ie geno ype and ma e nal dep essi e symp oms we e associa ed wi h a ela i ely g ea e di icul y disengaging a en ion om ea ul acial exp essions. No associa ions we e ound wi h in an s’ empe amen al ai s. Conclusions: Al e a ions in in an s’ na u al a en ional bias owa d ea ul acial exp essions may eme ge p io o he mani es a ion o emo ional and social beha io s and p o ide a sensi i e ma ke o ea ly emo ional de elopmen . Keywo ds: A en ion, acial exp ession, ea , yp ophan hyd oxylase 2 gene, in ancy, ma e nal s ess. In oduc ion The i s yea s o li e appea o cons i u e a pe iod o heigh ened plas ici y when an o ganism’s biological sys ems a e pa icula ly sensi i e o gene ic and expe ien ial in luences, and may unde go unc ional al e a ions ha a e ounda ional o indi idual emo ional ai s (He zman & Boyce, 2010). S udies examining how de elopmen al p ocesses du ing his pe iod a e in luenced by gene ic and en i onmen al ad e si y (e.g., Feldman e al., 2009; Holmboe e al., 2010) may hence p o ide impo an new ma ke s o iden i ying indi iduals a isk and o unde s anding he on ogene ic o igins o emo ional ulne abili y. One ac able, bu as ye li le examined, aspec o ea ly de elopmen ela es o unc ional eme gence o mechanisms ha media e a en ional igilance o biologically ele an s imuli (LeDoux, 2012). In human in an s, he i s signs o such igilance a e obse ed du ing he second hal o he i s yea o li e when in an s begin o disc imina e be ween acial exp essions and exhibi an a en ional bias owa d aces ha exp ess ea (Pel ola, Hie anen, Fo ssman, & Lepp€ anen, 2013). Simila igilance o dange - ale ing cues is obse ed in se e al species (Olsson & Phelps, 2007), sugges ing a conse ed mechanism ha may se e as an adap i e unc ion in de ec ing ha m ul s imuli. Ye , he e a e also indica ions ha gene ic and expe ien ial ac o s may exace ba e such species- ypical pe cep ual p edisposi ions (Shackman, Shackman, & Pollak, 2007), and ha a di icul y disengaging a en ion om h ea - ela ed acial exp essions is co ela ed wi h ai anxie y (Geo giou e al., 2005) and p edic s ulne abili y o anxie y in child en and adul s (Ba -Haim, Mo ag, & Glickman, 2011; Fox, Hane, & Pine, 2007; Pine, 2007). In he p esen s udy, we examined whe he he de elopmen o a en ion o emo ional cues is modula ed by es ablished gene ic and en i onmen al isk ac o s. Al hough a en ional biases canno be di ec ly linked wi h pa icula gene ic o en i on- men al ac o s, we easoned ha he apid de elop- men o hese p ocesses du ing he i s yea o li e may make hem suscep ible o modula o y e ec s by ac o s ha a e mo e nonspeci ic in na u e (c . Leona do & Hen, 2007). Con lic o in e es s a emen : No con lic s decla ed. ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. This is anopen access a icle unde he e ms o he C ea i e Commons A ibu ion-NonComme cial-NoDe i s License, which pe mi s use and dis ibu ion in any medium, p o ided he o iginal wo k is p ope ly ci ed, he use is non-comme cial and no modi ica ions o adap a ions a e made. Jou nal o Child Psychology and Psychia y 55:7 (2014), pp 793–801 doi:10.1111/jcpp.12181 The p ima y ocus o ou gene ic analyses was a common G o T base subs i u ion ( s4570625) in he p omo e egion o a gene ha encodes yp ophan hyd oxylase-2 (TPH2), a b ain-speci ic enzyme o se o onin syn hesis. The T-allele o his single nucleo ide polymo phism (SNP) has been associa ed wi h al e ed TPH2 mRNA exp ession (Chen, Vallende , & Mille , 2008), enhanced amygdala and co ical esponsi eness o emo ional cues (B own e al., 2005; He mann e al., 2007), educed a en- ion con ol (S obel e al., 2007), and inc eased suscep ibili y o dep essi e diso de s (Gao e al., 2012). The possibili y ha hese gene ic e ec s a e es ablished h ough al e ed neu ode elopmen is suppo ed by e idence showing ha TPH2 exp ession commences e y ea ly in he de eloping b ain (emb y- onic day 11 in mice, Waide , A a agi, Gu knech , & Lesch, 2011) and he esul s o ou p e ious s udy showing ha he THP2 -703 T-ca ie geno ype (i.e., T-ca ie s, including bo h TT homozygo es and he - e ozygo es) a e associa ed wi h inc eased di icul y disengaging a en ion om happy and ea ul exp es- sions in 7-mon h-old in an s (Lepp€ anen e al., 2011). Ou goal in he p esen s udy was o examine he eplicabili y o hese esul s in a new sample o in an s. A second gene ic a ia ion examined in he p esen s udy was a polymo phism in he se o onin au o e- cep o 1A gene (HTR1A SNP s6295 -1019 G/C). The C/C geno ype o his SNP has been associa ed wi h inc eased se o one gic one and heigh ened amyg- dala eac i i y o acial exp essions (Fak a e al., 2009). In ou p e ious s udy (Lepp€ anen e al., 2011), no signi ican associa ion o his SNP on a en ion disengagemen in in an s was ound, bu he esul s showed a end-le el associa ion be ween he C/C geno ype and inc eased di icul y disengaging. To examine whe he he p edic ed e ec s o he TPH2 SNP s4570625 (-703 G/T) geno ype and in an s’ a en ional bias o ea ul acial exp essions a e dependen on ea ly ea ing condi ions, we exam- ined ma e nal s ess and dep essi e symp oms. Ma e nal s ess and dep ession ha e been associ- a ed wi h ad e se changes in se e al aspec s o child de elopmen (Bo ns ein, A e be y, Mash, & Manian, 2011; Feldman e al., 2009), wi h some s udies (Bagne , Pe i , Lewinsohn, & Seeley, 2010), al hough no all (Naicke , Wickham, & Colman, 2012), sugges ing ha dep essi e symp oms du ing p egnancy and i s pos na al mon hs a e especially s ongly associa ed wi h ad e se ou comes. I is no ewo hy ha s ess and dep essi e symp oms, as assessed by ypical sel - epo ques ionnai es, a e likely o be linked wi h se e al changes in he child’s gene ic p edisposi ions as well as p e- and pos na al en i onmen (e.g., Feldman e al., 2009). Ou s udy was no aimed a easing apa hese ac o s bu a he o examine whe he ma e nal s ess as an es ablished isk condi ion is linked wi h an ea ly-eme ging and heigh ened a en ion bias owa d signals o nega i e emo ion. Me hods Pa icipan s The p esen analyses included da a om wo ongoing longi udinal s udies. Mos o he pa icipan s in he wo s udies came om u ban, middle-class amilies o Finnish o igin. The i s sample (Coho 1) consis ed o 7-mon h-old in an s (n=66) om a s udy ha was s a ed in Oc obe 2007 and consis ed o labo a o y assessmen s a 7, 24, and 48 mon hs o age. The me hodology and esul s o he 7-mon h assessmen we e epo ed in a p e ious publica ion (Lepp€ anen e al., 2011) and we e used he e in pooled analyses o TPH2 SNP s4570625 e ec s ac oss wo samples. The second sample (Coho 2) consis ed o 5–7-mon h-old in an s om a s udy ha began in Ap il 2012 and include labo a o y assessmen s a 5, 7, 12, 24, and 48 mon hs o age. DNA samples we e a ailable om 73 ull- e m (≥37 weeks) in an s (35 emales) wi h sco able a en ion assessmen s a 5 (M=151.58 days, SD =3.28 days) and 7 (M=213.44 days, SD =4.61 days) mon hs o age. Ma e nal s ess and dep ession da a we e a ailable o 87 in an s wi h sco able a en ion da a a 5 (M=151.84 days, SD =3.26 days) and 7 (M=213.13 days, SD =3.23 days) mon hs o age. Addi ional in an s en olled in he s udy we e excluded om one o he analyses because o missing DNA (n=20) o ques ionnai e da a (n=5), and bo h analyses because o p ema u e bi h (n=1), ussiness (n=6) o echnical p oblems du ing es ing (n=6), expe imen e ’s e o (n=1), d opping ou (n=1), o p o iding insu icien gaze da a (<50% ials in one o se e al condi ions a one o bo h ime poin s; n=12). Gene ic and s ess- ela ed da a om Coho 2 ha e no been epo ed p e iously, bu he a en ion da a o subg oup o in an s in his sample we e used in a p e ious epo by Pel ola e al. (2013). E hical pe mission o he s udy was ob ained om he E hical Commi ee o Tampe e Uni e si y Hospi al and a w i en in o med consen was gi en by he pa en s o he pa icipan s be o e he s a o he s udy. DNA ex ac ion and geno yping A olume o 3.0 ml EDTA-whole blood was aken om he pa icipan s when hey we e 7 mon hs by an expe ienced labo a o y nu se and s o ed a 20°C. DNA was ex ac ed by using QIAamp*DNA Blood Miniki and BioRobo s*M48 (Qiagen, Hilden, Ge many). Consis en wi h he analyses conduc ed o Coho 1 (Lepp€ anen e al., 2011), he i s DNA analyses o Coho 2 ocused on single nucleo ide polymo phisms (SNPs) in he TPH2 ( s4570625) and HTR1A -1019 ( s6295) genes. Geno yping was pe o med by using Taqman*SNP Geno yping Assays and ABI P ism 7900HT Sequence De ec ion Sys em (Applied Biosys ems, Fos e Ci y, CA, USA) wi h he ollowing assays: C_226207_10 ( s4570625) and C_11904666_10 ( s6295). No disc epancies we e ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. 794 Linda Fo ssman e al. J Child Psychol Psychia 2014; 55(7): 793–801 de ec ed in he geno yping esul s o duplica e sam- ples. Ma e nal s ess and dep essi e symp oms Assessmen o s ess ul li e e en s and ma e nal dep essi e symp oms was pe o med h ough ques ionnai es. The child’s biological mo he o bo h pa en s we e epo ed as he p ima y ca egi e o he child o all excep wo dyads wi h a he as he p ima y ca egi e . The mo he s we e gi en he ques ionnai es a e he 7-mon h-labo a o y isi , we e asked o ill hem ou wi hin 2 weeks om he isi , and e u n hem o he labo a o y in a p epaid en elope. The ecen li e e en s ques ionnai e (based on B ugha, Bebing on, Tennan , & Hu y, 1985) consis ed o 18 s ess ul li e e en s and one open- esponse i em (a ull desc ip ion o he e en s is p o ided in Table S3). The mo he s we e asked o ick a ‘yes’ box i he e en had occu ed du ing he pas 12 mon hs, and ick a ‘s ill a ec s me’ box i he e en was s ill ha ing an e ec on hei li e. The sum o all ‘yes’ esponses and ‘s ill a ec s me’ esponses (excep esponses o he i em ela ed o child bi h) we e calcula ed sepa a ely and we e used as measu es o ma e nal s ess. To assess ma e nal pos na al dep essi e symp oms, mo he s illed ou he 10-i em Edinbu gh Pos na al Dep ession Scale (EPDS, Re . Cox, Holden, & Sago sky, 1987). The sum sco es om he EPDS we e used o index dep essi e symp oms, wi h highe alues indica ing highe le els o dep essi e symp oms. Assessmen o a en ion o acial exp essions A de ailed desc ip ion o he pa adigm ha was used o assess a en ion disengagemen can be ound in Pel ola e al. (2013). Du ing he labo a o y assess- men a 5 and 7 mon hs o age, he in an s sa on hei pa en s lap a a ~60-cm iewing dis ance in on o a 23-inch moni o ha was pa o a co neal- e lec ion eye- acke (Tobii TX300; Tobii Technology, S ockholm, Sweden). E e y expe imen- al ial was p eceded by a p esen a ion o an a en ion-g abbing s imulus (a ed ci cle ha expanded om 0.4° o 4.3°in a con inuous ashion) o a ac he in an ’s a en ion o he cen e o he sc een. The expe imen e moni o ed he in an ’s gaze di ec ion h ough a hidden ideo came a, and when he child was a ending o he ed ci cle, p essed a bu on o s a he ial. On each ial, he in an s we e p esen ed wi h a con ol s imulus (i.e., a sc ambled ace) o one o h ee di e en acial exp essions (i.e., neu al, happy, o ea ul acial exp ession) on he cen e o he sc een. A e 1000 ms, he cen al s imulus was lanked by a la e al s imulus (a black-and-whi e pic u e o e i- cally a anged ci cles o a checke boa d pa e n) 13.6°equip obably on he le o igh o 3000 ms, as shown in Figu e 1. The p esen analyses a e based on he in an s’ a en ion disengagemen om he cen al s imuli du ing he i s pa o he es ing session (24 ials; 6 ials/condi ion). The second pa o s imulus p esen a ion was added o he pu poses o EEG measu emen and will be epo ed in a sepa a e publica ion. Iden ical o Coho 1, ideo eco dings o each in an ’s beha io in Coho 2 we e coded ame-by-- ame by an obse e who was blind o he s imulus condi ion. In cases whe e he ideo eco ding was missing o in alid (e.g., ideo came a did no eco d due o echnical e o s, ideo eco ding showed a de ia ion in ame a e, o he child was ou o iew), da a poin s we e eplaced wi h gaze da a om he eye- acke eco ding. A ial was conside ed in alid i he in an did no look a he cen al s imulus o a leas 75% du ing he ini ial 1000 ms p esen a ion, i he in an made an an icipa o y eye mo emen (i.e., eye mo emen commenced <160 ms a e he onse o he la e al s imulus), o an eye mo emen owa d an inco ec loca ion (i.e., no owa d he la e al s imulus). O he sco able ials, he numbe o missing a en ion shi s (i.e., no eye mo emen owa d he la e al s imulus du ing a ime window om 160 o 1000 ms a e he onse o he la e al s imulus) we e calcula ed and se ed as he depen- den a iable in he s udy. 1000 ms 3000 ms Figu e 1 The sequence o e en s and s imuli in he pa adigm used o assess in an s’ a en ion. The p opo ion o missing saccades om he cen al s imulus o he la e al s imulus was used as a measu e o a en ion disengagemen ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. doi:10.1111/jcpp.12181 TPH2 gene and li e s ess a e associa ed wi h a en ion o ea 795 Assessmen o empe amen Pa en s a ed hei child’s empe amen using he In an Beha io Ques ionnai e (Ro hba , 1981) when hei child was 7 mon hs. S a is ical analysis Ou i s analysis was aimed a examining whe he he p e iously epo ed associa ion be ween he T-allele o he TPH2 SNP s4570625 (-703 G/T) and a en ion disengagemen (numbe o missing a en ion shi s) was eplica ed in Coho 2. As in he p e ious s udy (Lepp€ anen e al., 2011), a dominan e ec o he T-allele was es ed by di iding he pa icipan s in o G/G homozygo es and T-ca ie s, and by conduc ing a Gene alized Es ima ing Equa- ions (a Poisson log linea model, SPSS 20) wi h Geno ype, Age, and Facial Exp ession as ac o s, he numbe o missing saccades as a dependen a iable (a Kolmogo o –Smi no es showed ha he numbe o missing saccades ollowed a Poisson dis ibu ion, ps>. 06), and a na u al log o he numbe o ials as an o se a iable. A co esponding analysis was conduc ed o examine he associa ion be ween HTR1A SNP s6295 (-1019 G/C geno ypes (G-ca i- e s s. C/C homozygo es) and a en ion disengage- men in Coho 2. Following he s a is ical app oach o he p e ious s udy (Lepp€ anen e al., 2011), an unadjus ed alpha le el o .05 was used o bo h s a is ical analyses. Second, we examined whe he he numbe o ma e nal s ess ul li e e en s we e associa ed wi h he a en ional bias owa d acial exp essions. Da a on s ess ul li e e en s we e a ailable o Coho 2 only. In he p ima y analysis, a iables desc ibing he numbe o li e e en s and he numbe o li e e en s ‘s ill a ec ing’ he mo he we e di ided in o h ee g oups (i.e., g oups wi h 0, 1, and 2 o mo e li e e en s), and hei e ec s we e examined by Gene - alized Es ima ing Equa ions wi h Li e E en s, Age, and Facial Exp ession as ac o s in a Poisson log linea model and he numbe o missing saccades as a dependen a iable. In a seconda y co ela ion analysis (Spea man ho), in ended o supplemen he o iginal analysis by using he o iginal ung ouped a iables, we co ela ed he aw numbe o li e e en s wi h a a iable ha desc ibed a en ional bias owa d ea ul acial exp essions (i.e., he di e ence in he p opo ion o missing a en ions shi s o all non ea ul s imuli and ea ul acial exp essions). Thi d, we examined whe he he e ec s o TPH2 SNP s4570625 (-703 G/T) we e dependen on ea - ing condi ion. Fo his analysis, da a om Coho s 1 and 2 we e pooled o a ain a maximal sample size, and mo he ’s a ings o dep essi e symp oms ha we e a ailable om bo h coho s we e used as a a iable e lec ing he ea ly ea ing condi ions. A uni a ia e analysis o a iance (ANOVA) was pe o med wi h geno ype (be ween ac o ; GG: n=88 s. T-ca ie s: n=51) as an independen a iable, EPDS sco es as a con inuous co a ia e, and he ea bias sco e as a dependen a iable. Es ima es o e ec sizes o independen a iables (i.e., p opo ion o a iance explained by he a iable) a e gi en as pa ial g 2 as p o ided by SPSS. The ea bias sco es me he s anda d c i e ia o uni a ia e no mali y (skewness =.26; ku osis =.63). Finally, al hough he p ima y pu pose o his s udy was o examine whe he he selec ed gene ic and psychosocial ac o s we e associa ed wi h he ea ly de elopmen o a en ion disengagemen , co - esponding analyses we e also conduc ed o examine whe he pa allel associa ions a e obse ed in mea- su es o in an empe amen . Fo hese analyses, indi idual - es s we e used o compa e TPH2 SNP s4570625 (-703 G/T) geno ypes wi h espec o di e en empe amen al scales (Ro hba , 1981), and Pea son co ela ions we e calcula ed o examine he associa ions be ween ma e nal s ess ul li e e en s and in an empe amen dimensions in Coho 2, as well as ma e nal dep essi e symp oms and in an empe amen al ai s using pooled da a om Coho 1 and 2. Resul s TPH2 SNP s4570625 (-703 G/T) and a en ion disengagemen The allelic equencies o he TPH2 SNP s4570625 (-703 G/T) we e in Ha dy–Weinbe g equilib ium in he eplica ion sample, =.06, p=.79, and in an s in he G/G (n=45) and T-ca ie (n=28; G/T: n=26, T/T: n=2) g oups did no di e in gende a io, age, bi h-weigh , ma e nal dep essi e sco es, s ess ul li e e en s sco es, o numbe o sco able ials (Table S2). The mino allele equency o s4570625 in he p esen sample (.2055) was compa able o ha in gene al Finnish popula ion sampling (.1953, Rai ak- a i e al., 2008). A Gene alized Es ima ing Equa ions (Poisson) model wi h geno ype, acial exp ession, and age as ac o s e ealed no signi ican main e ec o TPH2 SNP s4570625 (-703 G/T) on numbe o missing a en ion shi s, p<.10, bu he e was a signi ican main e ec o acial exp ession, 2 =126.05, d =3, p<.001, and a signi ican in e - ac ion be ween geno ype and acial exp ession, 2 =10.89, d =1, p=.012. Inspec ion o he mean p opo ion o missing a en ion shi s in he wo geno ype g oups shows ha he wo geno ype g oups did no di e in o e all le els o missing a en ion shi s, bu he T-ca ie s displayed a la ge ela i e inc ease in numbe o missing saccades o ea ul acial exp essions compa ed o he o he condi ions (M inc ease =.14, SD =.15 in he GG g oup s. M inc ease =.23, SD =.16 in he T-ca ie g oup). Thus, bo h G/G homozygo es and T-ca ie s show he ypical pa e ns o inc eased numbe o missing a en ions shi s in he con ex ea ul acial ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. 796 Linda Fo ssman e al. J Child Psychol Psychia 2014; 55(7): 793–801 exp essions (see Figu e S1), bu he ypical pa e n o inc eased numbe o missing shi s o ea is signi - ican ly la ge in he T-ca ie s. The e was no in e ac- ion be ween TPH2 -703 geno ype, emo ion, and age, p>.13, indica ing ha he e ec s o geno ype g oup and acial exp ession we e no age- ela ed. Toge he , hese esul s show ha he ela i e inc ease in miss- ing a en ion shi s o happy exp essions in T-ca i- e s (Lepp€ anen e al., 2011) did no eplica e in he new sample, bu he inc ease in missing saccades o ea ul acial exp ession was consis en ly la ge in T-ca ie s ac oss he wo s udies. Figu e 2 illus a es he geno ype e ec on p opo ion o missing a en ion shi s om he neu al and a ec i ely salien s imuli in he o al sample (Coho s 1 and 2 combined). Co esponding analysis o he HTR1A SNP s 6295 (-1019 G/C) showed no di e ences be ween he G-ca ie s (n=55; C/G: n=34, G/G: n=21) and C/C (n=18) geno ypes on a en ion disengagemen . This SNP was le ou om all subsequen da a analyses in he s udy. Ma e nal s ess and a en ion disengagemen The mean numbe o s ess ul li e e en s epo ed by he mo he s was 1.4 ( ange 0–5), wi h 23 mo he s epo ing no li e e en s, 33 one e en , and 31 wo o mo e li e e en s. The mean numbe o e en s s ill a ec ing he mo he was 0.6 ( ange 0–3), wi h 54 mo he s epo ing no a ec ing e en s, 19 epo ing one s ill a ec ing e en , and 14 epo ing wo o mo e a ec ing e en s. S ess ul li e e en s we e no asso- cia ed wi h he numbe o sco able ials in a en ion ask, all ps>.10, indica ing ha ma e nal s ess was no ela ed o he success o quali y o he da a collec ion pe se. A Gene alized Es ima ing Equa ions (Poisson) model wi h he numbe o s ess ul li e e en s, acial exp ession, and age as ac o s e ealed no signi ican main e ec o li e e en s no a signi ican in e ac ion be ween he numbe o li e e en s and acial exp ession on missing a en ion shi s, p=.12. A simila analysis using he numbe o li e e en s ‘s ill a ec ing’ he mo he e ealed a signi - ican in e ac ion be ween he numbe o e en s and acial exp ession, 2 =18.82, d =6, p=.004. As shown in Figu e 3, inc eased numbe o li e e en s ‘s ill a ec ing’ he mo he was associa ed wi h a linea inc ease in missing a en ion in he ea ul acial exp ession condi ion (i.e., ela i ely la ge ea bias sco es). The in e ac ion be ween he numbe o e en s and acial exp ession was u he quali ied by an in e ac ion wi h age, 2 =12.9, d =6, p=.045. Sepa a e analysis (uni a ia e ANOVAs) showed ha he associa ion be ween li e e en s and ea bias sco es was weake a 5 mon hs (p=.16) han a 7 mon hs o age (p=.003). The associa ion be ween s ess ul li e e en s and in an s’ ea bias sco es was also seen in co ela- ions be ween he aw (ung ouped) numbe s o li e e en s and ea bias sco es. The epo ed numbe o s ess ul li e e en s and he numbe o e en s ‘s ill a ec ing’ we e bo h signi ican ly associa ed wi h he s eng h o he ea bias in in an s, Spea man’s ho .32 and .37, ps<.004. TPH2 SNP s4570625 (-703 G/T), ma e nal dep ession, and in an s’ ea bias To examine whe he he obse ed e ec o he TPH2 SNP s4570625 (-703 G/T) geno ype on missing a en ion shi s was dependen on exposu e o Condi ion Fea HappyNeu alNon- ace Missing a en ion shi s (p) 0.50 0.40 0.30 0.20 0.10 0.00 T-ca ie GG TPH2-703 TPH2 Geno ype T-ca ie GG Fea bias sco e 1.00 0.75 0.50 0.25 0.00 –0.25 131 173 151 Figu e 2 Le : Mean p opo ion o missing a en ion shi s in he o al sample (Coho 1 and 2) om he cen al o he la e al s imulus in he a en ion disengagemen ask g ouped by acial exp essions and TPH2 SNP s4570625 (-703 G/T; G/G n=88; T-ca ie n=51). E o ba s ep esen 1 s anda d e o . Righ : A boxplo showing he e ec o TPH2 SNP s4570625 (-703 G/T) geno ypes on a en ion bias owa d ea ul acial exp essions, calcula ed as he di e ence in missing saccades o non ea ul (i.e., con ol, neu al, and happy s imuli) and ea ul acial exp essions ( ea bias sco e) ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. doi:10.1111/jcpp.12181 TPH2 gene and li e s ess a e associa ed wi h a en ion o ea 797 ma e nal dep essi e symp oms, we ca ied ou a uni a ia e ANOVA wi h geno ype (be ween ac o ; GG homozygo es s. T-ca ie s) as an independen a - iable and ma e nal dep essi e sco e as a co a ia e using pooled da a om Coho 1 and 2. This analysis showed a signi ican main e ec o TPH2 SNP s4570625 (-703 G/T), F(1, 130) =7.0, p=.009, pa ial g 2 =.05, and ma e nal dep ession, F(1, 130) =8.3, p=.005, pa ial g 2 =.06, on he p opo ion o missing a en ion shi s o ea . These main e ec s we e quali ied by a signi ican in e ac ion be ween geno ype and ma e nal dep ession, F(1, 129) =8.5, p<.001, pa ial g 2 =.12. As shown in Figu e 4, T-ca ie s wi h mo he s who sco ed ela i ely highe on pos na al dep ession exhibi ed highes le els o missing a en ion shi s o ea ul acial exp ession. No signi ican main e ec o in e - ac ions we e ound o Coho (be ween ac o ; Coho 1 s. Coho 2, p- alues =.14–.92.) o any subs an ial change o he abo e desc ibed esul s, when he ANOVA was conduc ed wi h his ac o as an independen a iable. TPH2 SNP s4570625 (-703 G/T), ma e nal s ess, and in an empe amen No associa ion be ween geno ype and any o he empe amen al dimensions we e ound in pooled analysis o da a om Coho 1 and 2 (n=157), o be ween ma e nal s ess and in an empe amen in Coho 2 (n=117), ps>.05. Ma e nal dep es- si e symp oms we e co ela ed wi h in an empe - amen al o ien ing, dis ess o limi a ions, and nega i e a ec i i y, s .18–.23, ps<.05 (unco - ec ed), in pooled analysis o da a om Coho 1 and 2. Discussion The e has been inc easing ecen in e es in he possibili y ha a ia ions in he TPH2 gene may al e ea ly neu ode elopmen , pa ially con ingen on exposu e o ad e se expe iences. Consis en wi h his possibili y, ou esul s showed ha he T-ca ie geno ype o he TPH2 SNP s4570625 (-703 G/T) is associa ed wi h a simila end o ela i ely g ea e Condi ion Fea ulHappyNeu alNon- ace Missing a en ion shi s (p) 0.60 0.50 0.40 0.30 0.20 0.10 0.00 2 o mo e 1 0 S ess ul li e e en s "s ill a ec ing" S ess ull li e e en s "s ill a ec ing" 2 o mo e10 Fea bias sco e 0.60 0.40 0.20 0.00 –0.20 –0.40 Figu e 3 Le : Numbe o s ess ul li e e en s ‘s ill a ec ing’ he mo he and mean p opo ions o in an s’ missing saccades in di e en s imulus condi ions. Righ : A boxplo showing he associa ion be ween ‘s ill a ec ing’ s ess ul e en s and ea bias sco es Condi ion Fea HappyNeu alNon- ace Missing a en ion shi s (p) 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 T-ca ie & high EPDS T-ca ie & low EPDS GG & high EPDS GG & low EPDS G oup Figu e 4 Mean p opo ion o missing a en ion shi s in he o al sample (Coho 1 and 2) om he cen al s imulus (Non ace, Neu al, Happy, o Fea ul acial exp ession) o he la e al a ge g ouped by ma e nal dep ession sco e (Edinbu gh Pos na al Dep ession Scale; EPDS; Median-spli : high s. low) and TPH2 SNP s4570625 (-703 G/T) geno ypes (G/G and low EPDS n=42; G/ G and high EPDS n=42; T-ca ie and low EPDS n=34; T-ca ie and high EPDS n=15). E o ba s ep esen 1 s anda d e o ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. 798 Linda Fo ssman e al. J Child Psychol Psychia 2014; 55(7): 793–801 di icul y disengaging a en ion om ea ul acial exp essions in wo independen samples o 5- o 7-mon h-old in an s. We also ound ha his asso- cia ion was dependen on he in an ’s ea ing condi- ions in a p edic ed way, being mos p onounced in T-ca ie s whose mo he s epo ed highe le els o dep essi e symp oms. I con i med, he po en ial dominan associa ion o he T-allele o he TPH2 SNP s4570625 (-703 G/T) wi h inc eased se o onin syn- hesis and se o one gic one (Chen & Mille , 2012), and he dependence o TPH2 exp ession on ea ly expe iences (Ga dne e al., 2009) may p o ide a neu obiologically plausible mechanism o enhanced a en ion o ea . Indeed, a andomized con olled s udy in human adul s has shown ha inc eased yp ophan and se o one gic unc ion heigh ens pe cep ual sensi i i y o ea ul acial exp essions (A enbu ow e al., 2003). A p esen , his in e p e- a ion mus be wi h cau ion, howe e , gi en he absence o di ec e idence o he unc ional signi - icance o he TPH2 SNP s4570625 (-703 G/T). Fu he mo e, he p esen s udy ocused on a single polymo phism in he TPH2 gene while i is likely ha mul iple SNPs, genes, and neu o ansmi e sys ems a e in ol ed in ea ly a en ional de elopmen . In u u e esea ch, i will be impo an o examine whe he he obse ed gene ic associa ions we e a ec ed by gene ic ances y by analyzing ances y in o ma i e gene ic ma ke s and by eplica ing he s udy in samples om di e en ances ies. I should also be no ed ha he cu en analyses do no ule ou he possibili y ha he obse ed gene ic associa ions in he in an s may be pa ly a ibu - able o indi ec e ec s a ising om he ac ha he pa en s a e likely o sha e he same geno ype and associa ed beha io al ai s (such as heigh ened dep ession). S ess ul li e e en s (pa icula ly when pe cei ed as ‘s ill a ec ing’ he mo he ) we e also associa ed wi h a selec i e di icul y disengaging a en ion om ea ul acial exp essions in in an s. Besides sha ed gene ic in luences be ween he mo he and he child, he e a e se e al plausible expe ience- ela ed mech- anisms ha may media e he associa ion be ween ma e nal s ess and heigh ened in an a en ion o ea ul acial exp essions. Ma e nal s ess may pose a nonspeci ic isk condi ion ha is associa ed wi h a numbe o p e- and pos na al ad e se in luences (e.g., educed ma e nal sensi i i y) and co-occu ing condi ions, each o which may lead o exace ba ions in in an s s ess eac i i y and ela ed p ocesses, possibly also inc easing in an s a en i eness o social signals o ea (Loman & Gunna , 2010). A ela ed possibili y is ha expe ienced s ess inc eases ca egi e s’ p o ec i e and p ecau iona y beha io s (Hahn-Holb ook, Holb ook, & Hasel on, 2011). P esumably, one o he p ima y means by which ca egi e s exe such beha io s is by acial signaling o ea . I con inued o e ime, epea ed exposu e o such beha io s may inc ease bo h adul s’ and in an s’ sensi i i y o ea cues (see Shackman e al., 2007 o a simila explana ion o p esumably adap i e enhancemen in a en ion o ang y acial exp essions in mal ea ed child en). Ou esul s we e limi ed by he eliance on sel - epo s as he only sou ce o in o ma ion ega ding ma e nal s ess and dep essi e symp oms, as well as ma e nal epo s as he only sou ce o in o ma ion conce ning in an s’ empe amen . Ne e heless, ques ionnai es ha e been ex ensi ely used in de el- opmen al esea ch and we e conside ed o p o ide a su icien p oxy o ea ing condi ion and in an s’ empe amen o he pu poses o he p esen s udy. I is clea , howe e , ha u he esea ch wi h a b oade app oach o s udying he ole o in an s’ ea ing en i onmen o he de elopmen o a en ion o social signals, o example h ough di ec home obse a ions and he assessmen o mo he –child in e ac ion, du ing he i s mon hs o li e will be impo an o unco e ing he mechanisms ha unde lie he associa ion be ween s ess and heigh - ened ea bias. Also, i will be impo an o ule ou he possibili y ha ma e nal dep ession biases he a ings o in an empe amen by using mo e objec- i e measu es o empe amen . In summa y, he p esen s udy sugges s ha in an s’ na u al a en ional bias owa d ea ul acial exp essions may be exace ba ed by gene ic and psychosocial isk ac o s wi hin a ela i ely sho ime pe iod in ea ly de elopmen . These al e a ions in a en ional biases may p o ide a sensi i e and accessible ma ke o ea ly emo ional de elopmen , as al e a ions in a en ion we e obse ed wi hou any concomi an changes in mo he - epo ed empe a- men al ai s. Al hough we did no examine he long- e m unc ional signi icance o heigh ened ea bias in he cu en s udy (such analyses will be conduc ed when longi udinal assessmen s ha e been comple ed), i is in iguing o no e ha a weal h o ecen e idence sugges s ha exagge a ed a en- ion o h ea -ale ing cues may ep esen one o he key on ogene ic mechanisms ha ca alyzes he de elopmen o ea - ela ed empe amen al ai s and inc eases ulne abili y o anxie y diso de s (e.g., Ba -Haim e al., 2011). Na u ally, he in e ence o such causal associa ions be ween a en ion and emo ion equi es no only longi udinal da a bu , ul ima ely, con olled expe imen s wi h echniques o a enua e maladap i e a en ion biases (Ba - Haim e al., 2011). Recen s udies ha e shown ha eye- acking sys ems can be used o enhance a en- ion disengagemen in in an s (Wass, Po ayska- Poms a, & Johnson, 2011), bu i is no ye known whe he such e ec s ex end o in an s a en ion biases owa d h ea -ale ing cues. Suppo ing in o ma ion Addi ional Suppo ing In o ma ion may be ound in he online e sion o his a icle: ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. doi:10.1111/jcpp.12181 TPH2 gene and li e s ess a e associa ed wi h a en ion o ea 799 Figu e S1 Dis ibu ion o missing saccades o ea ul acial exp essions ac oss he wo samples. Table S1 Mean p opo ion o missing saccades in he a en ion disengagemen ask. Table S2 Demog aphic in o ma ion, sum sco es o ma e nal a ing on he Edinbu gh Pos na al Dep ession Scale. Table S3 Li e e en s ques ionnai e. Appendix S1. Regula o y a ian o he TPH2 gene and ea ly li e s ess a e associa ed wi h heigh ened a en- ion o social signals o ea in in an s. Acknowledgemen s This esea ch was suppo ed by g an s om he Acad- emy o Finland (#218284), he Eu opean Resea ch Council (# 283763), he Tampe e Uni e si y Hospi al Medical Funds (g an 9M048 o 9N035 o T.L), Finnish Founda ion o Ca dio ascula Resea ch, Tampe e Tube culosis Founda ion and Emil Aal onen Founda- ion. The au ho s g a e ully acknowledge he e o s o he amilies who pa icipa ed in he s udies. The au ho s decla e hey ha e no po en ial o compe ing con lic o in e es s. Co espondence Linda Fo ssman, Tampe e Cen e o Child Heal h Resea ch, School o Medicine, Uni e si y o Tampe e, FIN-33014 Tampe e, Finland; Email: linda. o ssman@ u a. i Key poin s •The ea ly s ages o human de elopmen a e highly sensi i e o gene ic and en i onmen al in luences, and unc ional adap a ions du ing his pe iod may be ounda ional o li e ime emo ional ai s. •Li le is known abou he mechanisms ha eme ge du ing his pe iod o media e long- e m ou comes. •Ou s udy shows ha in an s’ na u al a en ional bias owa d ea ul acial exp essions is heigh ened in associa ion wi h es ablished gene ic and psychosocial isk ac o s. •Heigh ened a en ional bias owa d ea may p o ide a sensi i e and accessible ma ke o he assessmen o ea ly de elopmen and po en ially also o unde s anding he mechanisms ha media e emo ional ulne abili y. Re e ences A enbu ow, M.J., Williams, C., Odon iadis, J., Reed, A., Powell, J., Cowen, P.J., & Ha me , C.J. (2003). Acu e adminis a ion o nu i ionally sou ced yp ophan inc eases ea ecogni ion. Psychopha macology (Be l),169, 104–107. Bagne , D.M., Pe i , J.W., Lewinsohn, P.M., & Seeley, J.R. (2010). E ec o ma e nal dep ession on child beha io : A sensi i e pe iod? Jou nal o he Ame ican Academy o Child and Adolescen Psychia y,49, 699–707. Ba -Haim, Y., Mo ag, I., & Glickman, S. (2011). T aining anxious child en o disengage a en ion om h ea : A andomized con olled ial. Jou nal o Child Psychology and Psychia y,52, 861–869. Bo ns ein, M.H., A e be y, M.E., Mash, C., & Manian, N. (2011). Disc imina ion o acial exp ession by 5-mon h-old in an s o nondep essed and clinically dep essed mo he s. In an Beha io and De elopmen ,34, 100–106. B own, S.M., Pee , E., Manuck, S.B., Williamson, D.E., Dahl, R.E., Fe ell, R.E., & Ha i i, A.R. (2005). A egula o y a ian o he human yp ophan hyd oxylase-2 gene biases amygdala eac i i y. Molecula Psychia y,10, 884–888. B ugha, T., Bebing on, P., Tennan , C., & Hu y, J. (1985). The lis o h ea ening expe iences: A subse o 12 li e e en s ca ego ies wi h conside able long- e m con ex ual h ea . Psychological Medicine,15, 189–194. Chen, G., & Mille , G.M. (2012). Ad ances in yp ophan hyd oxylase-2 gene exp ession egula ion: New insigh s in o se o onin-s ess in e ac ion and clinical implica ions. Ame ican Jou nal o Medical Gene ics Pa B: Neu opsychia ic Gene ics,159B, 152–171. Chen, G., Vallende , E., & Mille , G. (2008). Func ional cha ac e iza ion o he human TPH2 5′ egula o y egion: Un ansla ed egion and polymo phisms modula e gene exp ession in i o.Human Gene ics,122, 645–657. Cox, J.L., Holden, J.M., & Sago sky, R. (1987). De ec ion o pos na al dep ession. de elopmen o he 10-i em Edinbu gh pos na al dep ession scale. B i ish Jou nal o Psychia y,150, 782–786. Fak a, E., Hyde, L.W., Go ka, A., Fishe , P.M., Munoz, K.E., Kimak, M., ... & Ha i i, A.R. (2009). E ec s o HTR1A C (-1019)G on amygdala eac i i y and ai anxie y. A chi es o Gene al Psychia y,66,33–40. Feldman, R., G ana , A., Pa ien e, C., Kane y, H., Kuin , J., & Gilboa-Schech man, E. (2009). Ma e nal dep ession and anxie y ac oss he pos pa um yea and in an social engagemen , ea egula ion, and s ess eac i i y. Jou nal o he Ame ican Academy o Child and Adolescen Psychia y,48, 919–927. Fox, N.A., Hane, A.A., & Pine, D.S. (2007). Plas ici y o a ec i e neu oci cui y: How he en i onmen a ec s gene exp ession. Cu en Di ec ions in Psychological Science,16, 1–5. Gao, J., Pan, Z., Jiao, Z., Li, F., Zhao, G., Wei, Q., & E angelou, E. (2012). TPH2 gene polymo phisms and majo dep ession– a me a-analysis. PLoS ONE,7, e36721. Ga dne , K.L., Hale, M.W., Old ield, S., Ligh man, S.L., Plo sky, P.M., & Low y, C.A. (2009). Ad e se expe ience du ing ea ly li e and adul hood in e ac o ele a e ph2 mRNA exp ession in se o one gic neu ons wi hin he do sal aphe nucleus. Neu oscience,163, 991–1001. Geo giou, G.A., Bleakley, C., Haywa d, J., Russo, R., Du on, K., El i i, S., & Fox, E. (2005). Focusing on ea : A en ional disengagemen om emo ional aces. Visual Cogni ion,12, 145–158. Hahn-Holb ook, J., Holb ook, C., & Hasel on, M.G. (2011). Pa en al p ecau ion: Neu obiological means and adap i e ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. 800 Linda Fo ssman e al. J Child Psychol Psychia 2014; 55(7): 793–801 ends. Neu oscience and Biobeha io al Re iews,35, 1052– 1066. He mann, M.J., Hu e , T., Mulle , F., Muhlbe ge , A., Pauli, P., Rei , A., & Lesch, K. (2007). Addi i e e ec s o se o onin anspo e and yp ophan hyd oxylase-2 gene a ia ion on emo ional p ocessing. Ce eb al Co ex,17, 1160–1163. He zman, C., & Boyce, T. (2010). How expe ience ge s unde he skin o c ea e g adien s in de elopmen al heal h. Annual Re iew o Public Heal h,31, 329–347. Holmboe, K., Nemoda, Z., Fea on, R.M.P., Csib a, G., Sas a i-Szekely, M., & Johnson, M.H. (2010). Polymo phisms in dopamine sys em genes a e associa ed wi h indi idual di e ences in a en ion in in ancy.De elopmen al Psychology, 46,404–416. LeDoux, J. (2012). Re hinking he emo ional b ain. Neu on,73, 653–676. Leona do, E.D., & Hen, R. (2007). Anxie y as a de elopmen al diso de . Neu opsychopha macology,33, 134–140. Lepp€ anen, J.M., Pel ola, M.J., Puu a, K., M€ an ymaa, M., Mononen, N., & Leh im€ aki, T. (2011). Se o onin and ea ly cogni i e de elopmen : Va ia ion in he yp ophan hyd oxylase 2 gene is associa ed wi h isual a en ion in 7-mon h-old in an s. Jou nal o Child Psychology and Psychia y,52, 1144–1152. Loman, M.M., & Gunna , M.R. (2010). Ea ly expe ience and he de elopmen o s ess eac i i y and egula ion in child en. Neu oscience and Biobeha io al Re iews,34, 867–876. Naicke , K., Wickham, M., & Colman, I. (2012). Timing o i s exposu e o ma e nal dep ession and adolescen emo ional diso de in a na ional canadian coho . PLoS ONE,7, e33422. Olsson, A., & Phelps, E.A. (2007). Social lea ning o ea . Na u e Neu oscience,10, 1095–1102. Pel ola, M.J., Hie anen, J.K., Fo ssman, L., & Lepp€ anen, J.M. (2013). The eme gence and s abili y o he a en ional bias o ea ul aces in in ancy. In ancy,18, 905–926. Pine, D.S. (2007). Resea ch e iew: A neu oscience amewo k o pedia ic anxie y diso de s. Jou nal o Child Psychology and Psychia y,48, 631–648. Rai aka i, O.T., Juonala, M., R€ onnemaa, T., Kel ikangas-J€ a inen, L., R€ as€ anen, L., Pie ik€ ainen, M., ... & Viika i, J.S. (2008). Coho p o ile: The Ca dio ascula Risk in Young Finns S udy. In e na ional Jou nal o Epidemiology,37, 1220–1226. Ro hba , M.K. (1981). Measu emen o empe amen in in ancy. Child De elopmen ,52, 569–578. Shackman, J.E., Shackman, A.J., & Pollak, S.D. (2007). Physical abuse ampli ies a en ion o h ea and inc eases anxie y in child en. Emo ion,7, 838–852. S obel, A., D eisbach, G., M€ ulle , J., Goschke, T., B ocke, B., & Lesch, K. (2007). Gene ic a ia ion o se o onin unc ion and cogni i e con ol. Jou nal o Cogni i e Neu oscience,19, 1923–1931. Waide , J., A a agi, N., Gu knech , L., & Lesch, K. (2011). T yp ophan hyd oxylase-2 (TPH2) in diso de s o cogni i e con ol and emo ion egula ion: A pe spec i e. Psychoneu oendoc inology,36, 393–405. Wass, S., Po ayska-Poms a, K., & Johnson, M. (2011). T aining a en ional con ol in in ancy. Cu en Biology,21, 1543–1547. Accep ed o publica ion: 17 Oc obe 2013 Published online: 5 Decembe 2013 ©2013 The Au ho s. Jou nal o Child Psychology and Psychia y published by John Wiley & Sons L d on behal o Associa ion o Child and Adolescen Men al Heal h. doi:10.1111/jcpp.12181 TPH2 gene and li e s ess a e associa ed wi h a en ion o ea 801