Educational attainment of same-sex and opposite-sex dizygotic twins: An individual-level pooled study of 19 twin cohorts

Ho mones and Beha io 136 (2021) 105054 A ailable online 3 Sep embe 2021 0018-506X/© 2021 The Au ho s. Published by Else ie Inc. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/). Educa ional a ainmen o same-sex and opposi e-sex dizygo ic wins: An indi idual-le el pooled s udy o 19 win coho s Ka i Sil en oinen a , b , * , Leonie H. Bogl c , d , Aline Jelenko ic e , , Ee o Vuoksimaa c , An i La ala a , Weilong Li a , Qihua Tan g , Dong eng Zhang h , Zengchang Pang i , Juan R. O do˜ nana j , k , Juan F. S´ anchez-Rome a j , k , Lucia Colod o-Conde j , l , Gonneke Willemsen m , Meike Ba els m , Ca ha ina E.M. an Beijs e eld m , Es he Reba o n , Robin P. Co ley o , B ooke M. Huib eg se p , John L. Hoppe q , , Jessica Tyle q , Glen E. Duncan s , Ded a Buchwald s , Judy L. Silbe g , He mine H. Maes u , Ch is ian Kandle , Wendy Cozen w , x , Amie E. Hwang w , Thomas M. Mack w , x , T acy L. Nelson y , Kei h E. Whi ield z , Emanuela Medda aa , Lo enza Nis ic` o aa , Vi gilia Toccaceli aa , Robe F. K uege ab , Ma McGue ab , Shandell Pahlen ac , Nicholas G. Ma in ad , Sa ah E. Medland ad , G an W. Mon gome y ad , Kauko Heikkil¨ a , Ca he ine A. De om ae , a , Robe F. Vlie inck ae , Ru h J.F. Loos ag , Pa ik K.E. Magnusson ah , Nancy L. Pede sen ah , Anna K. Dahl Aslan ai , ah , aj , Ma hew Ho op ak , A hula Suma hipala al , am , F uhling Rijsdijk an , Sisi a H. Si ibaddana al , ao , Richa d J. Rose ap , Tho kild I.A. Sø ensen aq , a , Do e I. Boomsma m , Jaakko Kap io , c a Depa men o Social Resea ch, Uni e si y o Helsinki, Helsinki, Finland b Osaka Uni e si y G adua e School o Medicine, Osaka Uni e si y, Osaka, Japan c Ins i u e o Molecula Medicine FIMM, Helsinki, Finland d Depa men o Epidemiology, Cen e o Public Heal h, Medical Uni e si y o Vienna, Vienna, Aus ia e Depa men o Physiology, Uni e si y o he Basque Coun y (UPV/EHU), Bilbao, Spain Depa men o Public Heal h, Uni e si y o Helsinki, Helsinki, Finland g Epidemiology, Bios a is ics and Biodemog aphy, Depa men o Public Heal h, Uni e si y o Sou he n Denma k, Odense, Denma k h Depa men o Public Heal h, Qingdao Uni e si y Medical College, Qingdao, China i Depa men o Noncommunicable Diseases P e en ion, Qingdao Cen e s o Disease Con ol and P e en ion, Qingdao, China j Depa men o Human Ana omy and Psychobiology, Uni e si y o Mu cia, Mu cia, Spain k IMIB-A ixaca, Mu cia, Spain l QIMR Be gho e Medical Resea ch Ins i u e, B isbane, Aus alia m Depa men o Biological Psychology, VU Uni e si y Ams e dam, Ams e dam, Ne he lands n Depa men o Gene ics, Physical An h opology and Animal Physiology, Uni e si y o he Basque Coun y (UPV/EHU), Bilbao, Spain o Ins i u e o Beha io al Gene ics, Uni e si y o Colo ado, Boulde , CO, USA p Ins i u e o Beha io al Science, Uni e si y o Colo ado, Boulde , CO, USA q Twin Resea ch Aus alia, Cen e o Epidemiology and Bios a is ics, The Uni e si y o Melbou ne, Melbou ne, Vic o ia, Aus alia Depa men o Epidemiology, School o Public Heal h, Seoul Na ional Uni e si y, Seoul, Republic o Ko ea s Washing on S a e Twin Regis y, Washing on S a e Uni e si y - Heal h Sciences Spokane, Spokane, WA, USA Depa men o Human and Molecula Gene ics, Vi ginia Ins i u e o Psychia ic and Beha io al Gene ics, Vi ginia Commonweal h Uni e si y, Richmond, VA, USA u Depa men o Human and Molecula Gene ics, Psychia y & Massey Cance Cen e , Vi ginia Commonweal h Uni e si y, Richmond, VA, USA Depa men o Psychology, Uni e si y o B emen, B emen, Ge many w Depa men o P e en i e Medicine, Keck School o Medicine o USC, Uni e si y o Sou he n Cali o nia, Los Angeles, CA, USA x USC No is Comp ehensi e Cance Cen e , Los Angeles, CA, USA y Depa men o Heal h and Exe cise Sciences and Colo ado School o Public Heal h, Colo ado S a e Uni e si y, USA z Psychology Depa men , Uni e si y o Ne ada Las Vegas, Ne ada, USA aa Is i u o Supe io e di Sani ` a Cen e o Beha iou al Sciences and Men al Heal h, Rome, I aly ab Depa men o Psychology, Uni e si y o Minneso a, Minneapolis, MN, USA ac Depa men o Psychology, Uni e si y o Cali o nia, Ri e side, Ri e side, CA, USA ad Gene ic Epidemiology Depa men , QIMR Be gho e Medical Resea ch Ins i u e, B isbane, Aus alia ae Cen e o Human Gene ics, Uni e si y Hospi als Leu en, Leu en, Belgium * Co esponding au ho a : Uni e si y o Helsinki, Facul y o Social Sciences, P.O. Box 18, FIN-00014, Finland. E-mail add ess: [email p o ec ed] (K. Sil en oinen). Con en s lis s a ailable a ScienceDi ec Ho mones and Beha io jou nal homepage: www.else ie .com/loca e/yhbeh h ps://doi.o g/10.1016/j.yhbeh.2021.105054 Recei ed 15 Ma ch 2021; Recei ed in e ised o m 16 Augus 2021; Accep ed 17 Augus 2021 Ho mones and Beha io 136 (2021) 105054 2 a Depa men o Obs e ics and Gynaecology, Ghen Uni e si y Hospi als, Ghen , Belgium ag The Cha les B on man Ins i u e o Pe sonalized Medicine, The Mindich Child Heal h and De elopmen Ins i u e, Icahn School o Medicine a Moun Sinai, New Yo k, NY, USA ah Depa men o Medical Epidemiology and Bios a is ics, Ka olinska Ins i u e , S ockholm, Sweden ai School o Heal h Sciences, Uni e si y o Sk¨ o de, Sk¨ o de, Sweden aj Ins i u e o Ge on ology and Aging Resea ch Ne wo k–J¨ onk¨ oping (ARN-J), School o Heal h and Wel a e J¨ onk¨ oping Uni e si y, J¨ onk¨ oping, Sweden ak NIHR Men al Heal h Biomedical Resea ch Cen e, Sou h London and Maudsley NHS Founda ion T us and, Ins i u e o Psychia y Psychology and Neu oscience, King's College London, London, UK al Ins i u e o Resea ch & De elopmen , Ba a amulla, S i Lanka am Resea ch Ins i u e o P ima y Ca e and Heal h Sciences, School o P ima y Ca e Resea ch (SPCR), Facul y o Heal h, Keele Uni e si y, S a o dshi e, UK an King's College London, MRC Social, Gene ic & De elopmen al Psychia y Cen e, Ins i u e o Psychia y, Psychology & Neu oscience, London, UK ao Facul y o Medicine & Allied Sciences, Raja a a Uni e si y o S i Lanka Saliyapu a, S i Lanka ap Depa men o Psychological & B ain Sciences, Indiana Uni e si y, Blooming on, IN, USA aq No o No disk Founda ion Cen e o Basic Me abolic Resea ch, Facul y o Heal h and Medical Sciences, Uni e si y o Copenhagen, Copenhagen, Denma k a Depa men o Public Heal h (Sec ion o Epidemiology), Facul y o Heal h and Medical Sciences, Uni e si y o Copenhagen, Copenhagen, Denma k ARTICLE INFO Keywo ds: Educa ion Twins Twin es os e one ans e hypo hesis Tes os e one exposu e ABSTRACT Compa ing wins om same- and opposi e-sex pai s can p o ide in o ma ion on po en ial sex di e ences in a a ie y o ou comes, including socioeconomic- ela ed ou comes such as educa ional a ainmen . I has been sugges ed ha his design can be applied o examine he pu a i e ole o in au e ine exposu e o es os e one o educa ional a ainmen , bu he e idence is s ill dispu ed. Thus, we es ablished an in e na ional da abase o win da a om 11 coun ies wi h 88,290 indi idual dizygo ic wins bo n o e 100 yea s and es ed o di e ences be ween wins om same- and opposi e-sex dizygo ic pai s in educa ional a ainmen . E ec sizes wi h 95% con idence in e als (CI) we e es ima ed by linea eg ession models a e adjus ing o bi h yea and win s udy coho . In con as o he hypo hesis, no di e ence was ound in women (β = − 0.05 educa ional yea s, 95% CI −0.11, 0.02). Howe e , men wi h a same-sex co- win we e sligh ly mo e educa ed han men ha ing an opposi e- sex co- win (β =0.14 educa ional yea s, 95% CI 0.07, 0.21). No consis en di e ences in e ec sizes we e ound be ween indi idual win s udy coho s ep esen ing Eu ope, he USA, and Aus alia o o e he coho s bo n du ing he 20 h cen u y, du ing which pe iod he sex di e ences in educa ion e e sed a o ing women in he la es bi h coho s. Fu he , no in e ac ion was ound wi h ma e nal o pa e nal educa ion. Ou esul s con adic he hypo hesis ha he e would be di e ences in he in au e ine es os e one le els be ween same- sex and opposi e-sex emale wins a ec ing educa ion. Ou indings in men may poin o social dynamics wi hin same-sex win pai s ha may bene i men in hei educa ional ca ee s. 1. In oduc ion The e a e well-es ablished sex di e ences in educa ional a ainmen . These di e ences ha e changed o e he coho s bo n du ing he 20 h cen u y: educa ional le el was highe in men han in women in he coho s bo n be o e he Wo ld Wa II, bu he ea e he sex di e ence disappea ed and hen in he mos ecen coho s women end o be mo e educa ed han men (OECD, 2015). The win s udy design o e s a na u al expe imen , which can shed mo e ligh on he backg ound o hese sex di e ences in educa ional a ainmen . Di e ences in es os e one le els du ing e al li e can ha e pe manen e ec s on b ain de elopmen and beha io , which is demons a ed by mo e masculinized beha io o gi ls wi h congeni al ad enal hype plasia and also suppo ed by a la ge body o e idence based on expe imen al s udies o non-human mammals (Hines, 2020). Gonadal ho mone ans e be ween e uses has been sugges ed o lead o highe es os e one exposu e o emales wi h a male a he han a emale co- win (Ah en eld e al., 2020). This is known as he Twin Tes os e one T ans e (TTT) hypo hesis sugges ing ha com- pa isons o same-sex (SS) and opposi e-sex (OS) wins can in o m he e ec s o p ena al es os e one exposu e on psychological and beha - io al ai s, including educa ional a ainmen . Howe e , e en when he ho monal ans e be ween e uses has been shown in o he mammals, he di ec e idence in humans is s ill lacking (Tapp e al., 2011). Pos - na al ac o s, such as socializa ion, may also con ibu e o he di e - ences be ween SS and OS pai s and a e di icul o exclude in obse a ional s udies (Loehlin and Ma in, 2000). Some s udies o SS and OS wins ha e suppo ed he TTT hypo hesis o cogni ion, whe eas he e idence o o he ai s has been less consis en (Ah en eld e al., 2020). The e a e a ew p e ious s udies es ing he di e ences be ween OS and SS win pai s in educa ional a ainmen . A Danish s udy did no ind any di e ences be ween SS and OS wins in academic pe o mance a 15–16 yea s o age, while he expec ed sex di e ences a o ing males in ma hema ics and emales in languages we e ound (Ah en eld e al., 2015). A Swedish s udy ound ha bo h SS male and emale wins had, on a e age, sligh ly be e ac- ademic pe o mance a he age o 15–16 yea s and a highe p obabili y o uni e si y educa ion as compa ed o hei OS coun e pa s (Hje n e al., 2012). A s udy o No wegian emale wins ound ha hose ha ing a male co- win had a highe a e o ailing o g adua e om high school o comple ing college as compa ed o SS emale wins (Bü iko e e al., 2019). Thus, he p e ious e idence may sugges ha wins om SS pai s, a leas in emales, ha e some academic ad an age as compa ed o OS wins. A d awback o hese p e ious s udies is ha since hey a e based on egis e da a, hey we e no able o make he dis inc ion be ween same- sex dizygo ic (SSDZ) wins and monozygo ic (MZ) wins. This is a lim- i a ion because dizygo ic (DZ) wins ha e, on a e age, sligh ly lowe educa ion han MZ wins (Sil en oinen e al., 2017). Thus, pooling MZ and SSDZ wins may ha e biased he esul s o such s udies. Fu he , all p e ious s udies ep esen ed No dic coun ies and included wins bo n in 1967 o la e . Thus, i is no known how uni e sal hese indings a e and whe he he changing sex di e ences in educa ion du ing he 20 h cen u y ha e a ec ed hese di e ences. To analyze his ques ion in- dep h, we conduc ed a la ge in e na ional s udy o wins. Ou da a co e 11 coun ies and bi h coho s o e 100 yea s as well as in o - ma ion on pa en al educa ion o sub-coho s, pe mi ing he analyses acco ding o mac o-le el socie al ac o s and amily en i onmen . We hypo hesize ha i he TTT hypo hesis is co ec o educa ion (i) he e should be di e ences be ween SSDZ and opposi e-sex dizygo ic (OSDZ) emale wins in educa ion; (ii) hese di e ences should ha e changed o e he bi h coho s when he sex di e ence in educa ion e e sed a o ing women; and (iii) he educa ion should be simila in SSDZ as OSDZ male wins. K. Sil en oinen e al. Ho mones and Beha io 136 (2021) 105054 3 2. Ma e ial and me hods The da a we e de i ed om he in e na ional CODATwins da abase a ge ed o ga he oge he all win coho s in he wo ld ha ing in o - ma ion on heigh and weigh (Sil en oinen e al., 2015; Sil en oinen e al., 2019). In his s udy, we analyzed hose coho s ha ing addi ional in o ma ion on one's own educa ion a 25 yea s o age o olde and including bo h SSDZ and OSDZ wins (a leas 50 wins in each g oup). The o iginal educa ional classi ica ions we e ans o med as educa ional yea s as desc ibed in de ail elsewhe e (Sil en oinen e al., 2017). Toge he , we had da a om 19 win coho s including 88,290 indi idual DZ wins (55% women and 39% OSDZ wins). In hese wins, we had in o ma ion on ma e nal educa ion o 24,417 wins and pa e nal edu- ca ion o 23,687 wins. Zygosi y was mainly assessed by alida ed ques ionnai e me hods, and in a mino i y o cases by gene ic es ing (Sil en oinen e al., 2015). The da a we e analyzed by linea eg ession models using S a a s a is ical package ( e sion 14.2) and using he clus e op ion o adjus o lack o s a is ical independence o he co- wins, when he p ima y sampling uni is he win pai (S a aCo p LP; College S a ion, TX, USA). This adjus men p o ides co ec s anda d e o s and con idence in- e als (CI) when wins a e analyzed as indi iduals. We adjus ed he esul s o bi h yea and win s udy coho o ake in o accoun he secula inc ease in educa ional le el as well as di e ences be ween educa ional le el and he educa ional classi ica ions used in each win s udy coho . To analyze he he e ogenei y be ween he win s udy co- ho s in e ec sizes, we conduc ed a me a-analysis using a maximum likelihood andom-e ec s es ima o wi h he “me an” command in S a a. The e ec sizes in win s udy coho s we e adjus ed o bi h yea . The he e ogenei y was measu ed by I 2 es ima ing he p opo ion o e - ec size a ia ion due o be ween-coho a ia ion. The alues o 25–49% ha e been sugges ed o indica e low, 50–75% mode a e and >75% high he e ogenei y (Higgins e al., 2003). 3. Resul s Table 1 shows how he di e ence in educa ional a ainmen be ween women and men measu ed as yea s o educa ion has changed o e he bi h coho s du ing he 20 h cen u y. In he coho s bo n be ween 1910 and 1939, men had sligh ly highe educa ion han women. A e ha he sex di e ences e e sed and in he coho s bo n a e 1940 women had highe educa ion han men. The la ges di e ence was ound in he la es bi h coho whe e women had, on a e age, 0.59 yea s highe educa ion compa ed o men. The s anda d de ia ions we e somewha g ea e in he ea lie compa ed o he la e bi h coho s. When analyzed in mo e de ail, his was because o a la ge p opo ion o hose ha ing e y limi ed educa ion in he ea lie bi h coho s (da a no shown bu a e a ailable om he co esponding au ho ). We hen analyzed he e ec sizes o educa ional a ainmen among SSDZ wins as compa ed o OSDZ wins (Table 2). In he pooled da a, no di e ence be ween SSDZ and OSDZ wins in women was ound (−0.05 educa ional yea s; 95% CI −0.11, 0.02). Howe e , men om SSDZ pai s had sligh ly g ea e educa ional a ainmen han OSDZ men (0.14 educa ional yea s; 95% CI 0.07, 0.21). The in e ac ion be ween own sex and he sex o he co- win was s a is ically signi ican (p <0.0001). When s a i ied by bi h coho s, di e ences be ween women om SSDZ and OSDZ pai s we e inconsis en , ha ing ei he a posi i e o a nega i e sign. In con as , he poin es ima es in men (excep o one coho ) showed an educa ional ad an age o SSDZ wins as compa ed o OSDZ wins. In bo h men and women, only one o hese bi h-coho s a i ied es ima es was s a is ically signi ican . Fig. 1 p esen s he o es plo o he e ec sizes by win s udy coho in women and Fig. 2 in men. The coho s we e o de ed so ha he i s nine coho s shown a e om Eu ope ollowed by six coho s om he USA, wo coho s om Aus alia and indi idual coho s om S i Lanka and China. The e was mode a e he e ogenei y be ween he win coho s in he e ec sizes: I 2 =71% o women and 65% o men. In women, he esul s we e mixed, such ha in 8 o 19 coho s he poin es ima es showed ad an age o SSDZ wins o e OSDZ wins. In men, 14 o 19 coho -speci ic poin es ima es showed educa ional ad an age o SSDZ wins o e OSDZ wins. In he only Eas -Asian coho om China, he zygosi y di e ence di e ed om he o he coho s, showing highe educa ion in SSDZ women and OSDZ men; howe e , he in e ac ion e ec be ween own sex and he sex o he co- win was no s a is ically signi ican in his coho (p =0.28). Finally, we es ed, in a sub-coho , whe he pa en al educa ion modi ies he di e ence in educa ional a ainmen be ween SSDZ and OSDZ wins. We ound no e idence o an in e ac ion e ec be ween he sex o he co- win and ma e nal o pa e nal educa ion, ei he in women (p =0.40 and 0.58, espec i ely) o in men (p =0.90 and 0.49, espec i ely). 4. Discussion In his la ge s udy o da a o igina ing om win coho s om 11 coun ies, we ound ha in women he sex o co- win was no associa ed wi h educa ional achie emen . Howe e , in men, educa ional achie e- men was sligh ly highe i hey had a male co- win han when hey had a emale co- win. This associa ion in men was sys ema ically ound in di e en Eu opean coun ies, he USA and Aus alia, and hus seems o be uni e sal wi hin he Wes e n cul u al con ex . The di e ence (0.14 educa ional yea s) is ela i ely small. In compa ison, he sex di e ence in educa ion (0.59 yea s, a la ges , ound in he la es coho s bo n in 1980–1989 yea s) is mo e han ou imes as la ge as his di e ence be ween same e sus opposi e sex male win pai s. Fu he , we did no ind e idence ha he associa ions ha e changed o e bi h coho s du ing he 20 h cen u y o ha i is modi ied by pa en al educa ion. The consis ency o esul s ac oss bi h coho s is especially in e es ing, because he sex di e ence a o ing men in he ea lie bi h coho s e e sed in he la e bi h coho s whe e women had highe educa ion. Table 1 The numbe o win indi iduals, he desc ip i e s a is ics and sex di e ences o educa ional yea s by bi h coho . Bi h coho Men Women Sex di e ence a N Mean SD N Mean SD β Con idence in e als LL UL 1900–1909 248 8.1 4.56 502 8.3 4.43 0.20 −0.67 1.06 1910–1919 1052 10.3 4.56 1656 9.9 4.27 −0.44 −0.84 −0.04 1920–1929 3141 10.4 4.55 3781 10.2 4.20 −0.22 −0.47 0.01 1930–1939 5579 10.6 4.59 5695 10.5 4.29 −0.07 −0.25 0.11 1940–1949 10,940 11.4 4.47 11,944 11.6 4.22 0.21 0.09 0.33 1950–1959 10,100 12.8 3.85 12,092 13.1 3.68 0.27 0.15 0.39 1960–1969 4142 14.0 2.88 5782 14.0 2.82 0.05 −0.07 0.17 1970–1979 3456 14.2 2.81 5086 14.7 2.69 0.56 0.43 0.68 1980–1989 1337 14.1 2.72 1757 14.7 2.65 0.59 0.40 0.78 a Educa ional yea s in women minus educa ional yea s in men. K. Sil en oinen e al. Ho mones and Beha io 136 (2021) 105054 4 In ou only Eas -Asian coho om China, he esul s di e ed om he o he coho s, and educa ion was highe in SSDZ women and in OSDZ men. This esul should be ea ed wi h cau ion since he in e ac ion be ween own sex and he sex o co- win was no s a is ically signi ican , bu i may sugges di e ences be ween cul u al con ex s. I is no e- wo hy, howe e , ha we did no ind any e idence o disc imina ion o gi ls in his Chinese coho o educa ion since educa ional le el was sligh ly highe in women (10.8 educa ional yea s) han in men (10.0 educa ional yea s). Ou esul s di e om p e ious s udies om Sweden and No way epo ing highe educa ion in SS emale wins as compa ed o OS emale wins (Bü iko e e al., 2019; Hje n e al., 2012). A ac o which may explain hese di e ences as compa ed o ou esul s is ha in hese p e ious s udies MZ wins we e pooled wi h SSDZ wins, and MZ wins ha e sligh ly highe educa ional a ainmen han DZ wins (Sil en oinen e al., 2017). The eason o his zygosi y di e ence is unknown, bu could be associa ed wi h he highe ma e nal age o DZ wins. This can a ec he educa ion o o sp ing because o he numbe o olde siblings, a amily s uc u e known o be associa ed wi h lowe educa ional a ainmen in younge siblings (Black e al., 2005; B oo h and Kee, 2009). Thus, he mix u e o SSDZ and MZ pai s in p e ious s udies may explain why di e ences ha e also been ound be ween SS and OS pai s in emales. The lack o di e ence in educa ional achie emen be ween SSDZ and OSDZ pai s in women is no compa ible wi h he TTT hypo hesis o educa ional achie emen . This nega i e esul aises he ques ion o how plausible he TTT hypo hesis is in gene al in humans. The mos e idence o he TTT hypo hesis in humans conce ns cogni ion, whe eas he e is li le suppo o he TTT hypo hesis o o he ai s (Ah en eld e al., 2020). Howe e , e en o spa ial abili y showing consis en sex di e - ences (Pe e s e al., 2006), he e idence o he TTT hypo hesis is inconsis en and he e a e s udies suppo ing (Heil e al., 2011; Vuok- simaa e al., 2010) and no suppo ing he expec ed di e ence be ween SS and OS emale wins (Toi ainen e al., 2018). I is also no ewo hy ha while he e is solid e idence o he TTT hypo hesis in li e -bea ing mammals, he di ec e idence o humans is lacking (Ryan and Van- denbe gh, 2002). The e a e majo physiological, molecula and s uc- u al di e ences o he placen a be ween human and oden Table 2 The di e ences o educa ional yea s o same-sex dizygo ic wins compa ed o opposi e-sex dizygo ic wins by sex and bi h coho . a Bi h coho Men Women p- alue o sex in e ac ion β Con idence in e als β Con idence in e als LL UL LL UL All 0.14 0.07 0.21 −0.05 −0.11 0.02 <0.0001 1900–1909 1.16 −0.77 3.10 0.90 −0.38 2.18 0.742 1910–1919 0.23 −0.40 0.85 −0.18 −0.65 0.29 0.710 1920–1929 0.09 −0.26 0.45 −0.12 −0.42 0.19 0.754 1930–1939 0.13 −0.12 0.37 −0.05 −0.29 0.18 0.886 1940–1949 0.27 0.13 0.41 0.13 0.00 0.26 0.135 1950–1959 −0.05 −0.20 0.09 −0.13 −0.25 −0.01 0.674 1960–1969 0.06 −0.11 0.23 0.02 −0.13 0.16 0.455 1970–1979 0.14 −0.06 0.34 −0.03 −0.19 0.13 0.099 1980–1989 0.17 −0.13 0.47 −0.01 −0.28 0.25 0.241 a Educa ional yea s o same-sex dizygo ic wins minus educa ional yea s o opposi e-sex dizygo ic wins. Fig. 1. The di e ence in educa ional yea s o same-sex dizygo ic wins as compa ed o opposi e-sex dizygo ic wins by win coho in women. K. Sil en oinen e al. Ho mones and Beha io 136 (2021) 105054 5 p egnancies, which make ex apola ion o esul s om oden s udies o humans unce ain (Schmid e al., 2015). Since he e idence is lacking whe he he e can e en be a ans usion o gonadal ho mones be ween human e uses, i is impossible o say whe he ou nega i e esul s sugges ha es os e one le els in he u e us do no a ec educa ion le el o whe he he e a e no any di e ences be ween SS and OS emale wins in hei in au e ine es os e one exposu e le els. The easons o ou indings ha educa ional le el was highe in SSDZ han in OSDZ men a e no clea . Low bi hweigh is ound o be associa ed wi h less educa ion (B eslau e al., 2004). Howe e , bi h- weigh is lowe in SSDZ han in OSDZ male wins and hus canno explain his associa ion (Jelenko ic e al., 2018). I is possible ha he easons may be ela ed o pos - a he han p ena al li e. The e is some e idence ha pa en s may inancially suppo he educa ion o hei sons mo e han hei daugh e s (Raley and Bianchi, 2006). Howe e , i his kind o sex disc imina ion exis s, i should a he bene i sons wi h a win sis e mo e han a win b o he . Finally, i is possible ha he social dynamics o SS male wins may acili a e hei educa ional ca ee . Fo ins ance, a male coun e pa in SS win pai s may se e as a model o as a compe i o , mo i a ing educa ional ad ancemen . Howe e , his is all specula i e and mo e esea ch on he in e nal dynamics wi hin win pai s is needed o shed mo e ligh on his inding. I is, howe e , no e- wo hy how uni e sal his e ec is, a leas wi hin he Wes e n social con ex . Ou da a ha e majo s eng hs bu also limi a ions. Ou main s eng h is ha we u ilized a e y la ge pooled da ase om 19 inde- penden ly collec ed win coho s ep esen ing 11 coun ies and bi h coho s o e 100 yea s. The need o he eplica ion o esul s and he p oblem o publica ion bias a e well- ecognized in psychology (La away e al., 2019), and hese p oblems may be especially impo an in his ype o s udy es ing a speci ic hypo hesis when he e can be a publi- ca ion bias agains esul s no suppo ing a p io i hypo hesis. Ou s udy can answe bo h p oblems by eplica ing he analyses in mul iple win coho s and publishing he es ima es i espec i e o e ec sizes, s a is- ical signi icance o whe he hey suppo he p e ious hypo hesis o no . I is also a clea ad an age ha we can dis inguish MZ pai s om SSDZ pai s, which may ha e biased he esul s ound in p e ious s udies on his opic. Howe e , o mos win pa icipan s in his s udy, he zygosi y was no con i med by DNA. Since educa ion is somewha highe in MZ as compa ed o DZ wins (Sil en oinen e al., 2017), possible misclassi ica ions o MZ and SSDZ wins may ha e con ibu ed o he highe educa ion in SSDZ wins as compa ed o OSDZ wins in men. Howe e , his canno explain ou main inding, i.e., he lack o di e - ence in educa ion be ween OSDZ and SSDZ wins in women. Ou main limi a ion is ha we only had in o ma ion on educa ional a ainmen and no academic pe o mance o IQ as a possible de e minan o edu- ca ion, which could help o s udy his e ec in mo e de ail. Addi ionally, ou da a a e based on su eys ins ead o egis e da a and a e hus p one o non- esponse bias. The p opo ion o OSDZ wins is gene ally some- wha lowe in win su eys han he p opo ion o SSDZ wins, indi- ca ing a lowe esponse a e (Sil en oinen e al., 2015). I is u he known ha low educa ion is ypically associa ed wi h lowe esponse a es (Reinikainen e al., 2018). Howe e , we assume ha his bias should a he inc ease he le el o educa ional a ainmen in OSDZ wins i hose ha ing lowe educa ion ha e disp opo ionally no pa icipa ed in he su eys. Mo e s udies a e also needed o in es iga e he social dynamics wi hin SS and OS win pai s, which may help o explain ou indings o men. Fu he , we had limi ed da a on non-Wes e n pop- ula ions. Mo e win s udies conduc ed in non-Wes e n socie ies a e needed o s udy whe he he esul s can be gene alized ou side he Wes e n cul u al con ex . In conclusion, ou esul s do no suppo he hypo hesis ha he e would be di e ences in he in au e ine es os e one le els be ween OSDZ and SSDZ emale wins a ec ing educa ion. Men ha ing a same sex co- win ha e sligh ly g ea e educa ional a ainmen han men wi h an opposi e sex co- win. This di e ence is uni e sal wi hin, a leas , he Wes e n cul u al con ex and is no modi ied by bi h coho o pa en al educa ion. Whe eas we canno exclude a p ena al biological explana- ion, we ind i mo e plausible ha pos na al socializa ion o male wins may c ea e he di e ences, and his hypo hesis should be conside ed in u u e esea ch. Fig. 2. The di e ence in educa ional yea s o same-sex dizygo ic wins as compa ed o opposi e-sex dizygo ic wins by win coho in men. K. Sil en oinen e al. Ho mones and Beha io 136 (2021) 105054 6 Acknowledgemen s This s udy was conduc ed wi hin he CODATwins p ojec . Suppo o collabo a o s: Since i s o igin he Eas Flande s P ospec i e Su ey has been pa ly suppo ed by g an s om he Fund o Scien i ic Resea ch, Flande s and Twins, a non-p o i Associa ion o Scien i ic Resea ch in Mul iple Bi hs (Belgium). Da a collec ion and analyses in Finnish win coho s ha e been suppo ed by ENGAGE – Eu opean Ne wo k o Gene ic and Genomic Epidemiology, FP7-HEALTH-F4- 2007, g an ag eemen numbe 201413, Na ional Ins i u e o Alcohol Abuse and Alcoholism (g an s AA-12502, AA-00145, and AA-09203 o R J Rose), he Academy o Finland Cen e o Excellence in Complex Dis- ease Gene ics (g an numbe s: 213506, 129680), and he Academy o Finland (g an s 100499, 205585, 118555, 141054, 265240, 263278 and 264146 o J Kap io). The Mu cia Twin Regis y is suppo ed by Fundaci´ on S´ eneca, Regional Agency o Science and Technology, Mu - cia, Spain (08633/PHCS/08, 15302/PHCS/10 & 19479/PI/14) and Minis y o Science and Inno a ion, Spain (PSI2009-11560, PSI2014- 56680-R and RTI2018-095185-B-I00, co- unded by Eu opean Regional De elopmen Fund (FEDER)). Ne he lands Twin Regis e acknowledges he Ne he lands O ganiza ion o Scien i ic Resea ch (NWO) and MagW/ZonMW g an s 904-61-090, 985-10-002, 912-10-020, 904-61- 193,480-04-004, 463-06-001, 451-04-034, 400-05-717, Addic ion- 31160008, Middelg oo -911-09-032, Spinozap emie 56-464-14192; VU Uni e si y's Ins i u e o Heal h and Ca e Resea ch (EMGO+); he Eu opean Resea ch Council (ERC - 230374), he A e a Ins i u e, Sioux Falls, Sou h Dako a (USA). Cali o nia Twin P og am was suppo ed by The Cali o nia Tobacco-Rela ed Disease Resea ch P og am (7RT-0134H, 8RT-0107H, 6RT-0354H) and he Na ional Ins i u es o Heal h (1R01ESO15150-01). The Ca olina A ican Ame ican Twin S udy o Aging (CAATSA) was unded by a g an om he Na ional Ins i u e on Aging (g an 1RO1-AG13662-01A2) o K. E. Whi ield. Colo ado Twin Regis y is unded byNIDA unded cen e g an DA011015, & Longi- i udinal Twin S udy HD10333; Au ho Huib eg se is suppo ed by 5T32DA017637 and 5T32AG052371 Washing on S a e Twin Regis y ( o me ly he Uni e si y o Washing on Twin Regis y) was suppo ed in pa by g an NIH RC2 HL103416 (D. Buchwald, PI). This esea ch was acili a ed h ough access o Twins Resea ch Aus alia, a na ional esou ce suppo ed by a Cen e o Resea ch Excellence G an (ID: 1079102), om he Na ional Heal h and Medical Resea ch Council. Re e ences Ah en eld , L., Pe e sen, I., Johnson, W., Ch is ensen, K., 2015. Academic pe o mance o opposi e-sex and same-sex wins in adolescence: a Danish na ional coho s udy. Ho m. Beha . 69, 123–131. Ah en eld , L.J., Ch is ensen, K., Segal, N.L., Hu , Y.M., 2020. Opposi e-sex and same-sex win s udies o physiological, cogni i e and beha io al ai s. Neu osci. Biobeha . Re . 108, 322–340. Black, S.E., De e eux, P.J., Sal anes, K.G., 2005. The mo e he me ie ? The e ec o amily size and bi h o de on child en's educa ion. Q. J. Econ. 120, 669–700. B eslau, N., Pane h, N.S., Lucia, V.C., 2004. The linge ing academic de ici s o low bi h weigh child en. Pedia ics 114, 1035–1040. B oo h, A.L., Kee, H.J., 2009. Bi h o de ma e s: he e ec o amily size and bi h o de on educa ional a ainmen . J. Popul. Econ. 22, 367–397. Bü iko e , A., Figlio, D.N., Ka bownik, K., Kuzawa, C.W., Sal anes, K.G., 2019. E idence ha p ena al es os e one ans e om male wins educes he e ili y and socioeconomic success o hei emale co- wins. P oc. Na l. Acad. Sci. U. S. A. 116, 6749–6753. Heil, M., Ka ˇ sek, M., Rolke, B., Bes e, C., Jansen, P., 2011. Men al o a ion in emale a e nal wins: e idence o in a-u e ine ho mone ans e ? Biol. Psychol. 86, 90–93. Higgins, J.P., Thompson, S.G., Deeks, J.J., Al man, D.G., 2003. Measu ing inconsis ency in me a-analyses. BMJ 327, 557–560. Hines, M., 2020. Human gende de elopmen . Neu osci. Biobeha . Re . 118, 89–96. Hje n, A., Ekeus, C., Rasmussen, F., Lindblad, F., 2012. Educa ional achie emen and oca ional ca ee in wins - a Swedish na ional coho s udy. Ac a Paedia . 101, 591–596. Jelenko ic, A., Sund, R., Yokoyama, Y., Hu , Y.M., Ullema , V., Almq is , C., Sil en oinen, K., 2018. Bi h size and ges a ional age in opposi e-sex wins as compa ed o same-sex wins: an indi idual-based pooled analysis o 21 coho s. Sci. Rep. 19 (6300). La away, S., Snyce ski, S., P adhan, S., Hui ema, B.E., 2019. An o e iew o scien i ic ep oducibili y: conside a ion o ele an issues o beha io science/analysis. Pe spec . Beha . Sci. 42, 33–57. Loehlin, J.C., Ma in, N.G., 2000. Dimensions o psychological masculini y- eminini y in adul wins om opposi e-sex and same-sex pai s. Beha . Gene . 30, 19–28. OECD, 2015. The ABC o Gende Equali y in Educa ion: Ap i ude, Beha iou , Con idence. PISA, OECD Publishing. h ps://doi.o g/10.1787/9789264229945-en. Pe e s, M., Lehmann, W., Takahi a, S., Takeuchi, Y., Jo dan, K., 2006. Men al o a ion es pe o mance in ou c oss-cul u al samples (n =3367): o e all sex di e ences and he ole o academic p og am in pe o mance. Co ex 42, 1005–1014. Raley, S., Bianchi, S., 2006. Sons, daugh e s, and amily p ocesses: does gende o child en ma e . Annu. Re . Sociol. 32, 401–421. Reinikainen, J., Tolonen, H., Bo odulin, K., H¨ a k¨ anen, T., Jousilah i, P., Ka anen, J., Va iainen, E., 2018. Pa icipa ion a es by educa ional le els ha e di e ged du ing 25 yea s in Finnish heal h examina ion su eys. Eu . J. Pub. Heal h 28, 237–243. Ryan, B.C., Vandenbe gh, J.G., 2002. In au e ine posi ion e ec s. Neu osci. Biobeha . Re . 26, 665–678. Schmid , A., Mo ales-P ie o, D.M., Pas uschek, J., F ¨ ohlich, K., Ma ke , U.R., 2015. Only humans ha e human placen as: molecula di e ences be ween mice and humans. J. Rep od. Immunol. 108, 65–71. Sil en oinen, K., Jelenko ic, A., Sund, R., Honda, C., Aal onen, S., Yokoyama, Y., Kap io, J., 2015. The CODATwins p ojec : he coho desc ip ion o COllabo a i e p ojec o De elopmen o An h opome ical measu es in wins o s udy mac o- en i onmen al a ia ion in gene ic and en i onmen al e ec s on an h opome ic ai s. Twin Res. Hum. Gene . 18, 348–360. Sil en oinen, K., Jelenko ic, A., La ala, A., Sund, R., Yokoyama, Y., Ullema , V., Kap io, J., 2017. Educa ion in wins and hei pa en s ac oss bi h coho s o e 100 yea s: an indi idual-le el pooled analysis o 42 win coho s. Twin Res. Hum. Gene . 20, 395–405. Sil en oinen, K., Jelenko ic, A., Yokoyama, Y., Sund, R., Sugawa a, M., Tanaka, M., Kap io, J., 2019. The CODATwins p ojec : he cu en s a us and ecen indings o COllabo a i e p ojec o De elopmen o An h opome ical measu es in wins. Twin Res. Hum. Gene . 20, 800–808. Tapp, A.L., Maybe y, M.T., Whi ehouse, A.J.O., 2011. E alua ing he win es os e one ans e hypo hesis: a e iew o he empi ical e idence. Ho m. Beha . 60, 713–722. Toi ainen, T., Pannini, G., Papageo giou, K.A., Malanchini, M., Rim eld, K., Shakesha , N., Ko as, Y., 2018. P ena al es os e one does no explain sex di e ences in spa ial abili y. Sci. Rep. 8 (13653). Vuoksimaa, E., Kap io, J., K emen, W.S., Hokkanen, L., Viken, R.J., Tuulio- Hen iksson, A., Rose, R.J., 2010. Ha ing a male co- win masculinizes men al o a ion pe o mance in emales. Psychol. Sci. 21, 1069–1071. K. Sil en oinen e al.