How Are Adolescents Sleeping? Adolescent Sleep Patterns and Sociodemographic Differences in 24 European and North American Countries

This is a sel -a chi ed e sion o an o iginal a icle. This e sion may di e om he o iginal in pagina ion and ypog aphic de ails. Au ho (s): Ti le: Yea : Ve sion: Copy igh : Righ s: Righ s u l: Please ci e he o iginal e sion: CC BY-NC-ND 4.0 h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ How A e Adolescen s Sleeping? Adolescen Sleep Pa e ns and Sociodemog aphic Di e ences in 24 Eu opean and No h Ame ican Coun ies © 2020 Else ie Published e sion Ga iepy, Gene ie e; Danna, So ia; Gobiņa, Inese; Rasmussen, Me e; Gaspa de Ma os, Ma ga ida; Tynjälä, Jo ma; Janssen, Ian; Kalman, Michal; Ville uša, Ani a; Husa o a, Daniela; B ooks, Fiona; Elga , F ank J.; Kla ina-Mak ecka, Sol i a; Šmigelskas, Kas y is; Gaspa , Tania; Schnoh , Ch is ina Ga iepy, G., Danna, S., Gobiņa, I., Rasmussen, M., Gaspa de Ma os, M., Tynjälä, J., Janssen, I., Kalman, M., Ville uša, A., Husa o a, D., B ooks, F., Elga , F. J., Kla ina-Mak ecka, S., Šmigelskas, K., Gaspa , T., & Schnoh , C. (2020). How A e Adolescen s Sleeping? Adolescen Sleep Pa e ns and Sociodemog aphic Di e ences in 24 Eu opean and No h Ame ican Coun ies. Jou nal o Adolescen Heal h, 66(6, Supplemen ), S81-S88. h ps://doi.o g/10.1016/j.jadoheal h.2020.03.013 2020 O iginal a icle How A e Adolescen s Sleeping? Adolescen Sleep Pa e ns and Sociodemog aphic Di e ences in 24 Eu opean and No h Ame ican Coun ies Gene ie e Ga iepy, Ph.D. a , * ,Sofia Danna, M.Sc. b , Inese Gobin¸ a, Ph.D. c , Me e Rasmussen, Ph.D. d , Ma ga ida Gaspa de Ma os, Ph.D. e , Jo ma Tynjälä, Ph.D. , Ian Janssen Ph.D. g , Michal Kalman Ph.D. h , Ani a Ville u sa, M.D. i , Daniela Husa o a, Ph.D. j , Fiona B ooks, Ph.D. k , F ank J. Elga , Ph.D. l , Sol i a Kla ina-Mak ecka M.Sc. m , Kas y is  Smigelskas, Ph.D. n , Tania Gaspa , Ph.D. o , and Ch is ina Schnoh , Ph.D. p a Depa men o Social and P e en i e Medicine, Uni e si y o Mon eal, Mon eal, Canada b Douglas Men al Heal h Uni e si y Ins i u e, Mon eal, Canada c Depa men o Public Heal h and Epidemiology, Ins i u e o Public Heal h, Riga S adins Uni e si y, Riga, La ia d Na ional Ins i u e o Public Heal h, Uni e si y o Sou he n Denma k, Copenhagen, Denma k e Ins i u e o En i onmen al Heal h, School o Human Kine ics, Uni e si y o Lisbon, Lisbon, Po ugal Facul y o Spo and Heal h Sciences, Uni e si y o Jy askyla, Jy askyla, Finland g Depa men o Public Heal h Sciences, School o Kinesiology and Heal h S udies, Queen's Uni e si y, Kings on, Canada h Depa men o Rec ea ion and Leisu e S udies, Facul y o Physical Cul u e, Palacky Uni e si y Olomouc, Olomouc, Czech Republic i Depa men o Public heal h and Epidemiology, Ins i u e o Public Heal h, Riga S adins Uni e si y, Riga, La ia j Depa men o Heal h Psychology, Facul y o Medicine, PJ Sa a ik Uni e si y, Kosice, Slo akia k O fice o he Depu y Vice-Chancello Resea ch, Uni e si y o Technology Sydney, Ul imo, Sydney, Aus alia l Ins i u e o Heal h and Social Policy, McGill Uni e si y, Mon eal, Canada m Depa men o Public Heal h and Epidemiology, Riga S adins Uni e si y, Riga, La ia n Heal h Resea ch Ins i u e, Li huanian Uni e si y o Heal h Sciences, Kaunas, Li huania o Psychology and Educa ional Sciences Ins i u e, Lusiada Uni e si y/CLISSIS, ISAMB/Lisbon Uni e si y, Lisbon, Po ugal p Depa men o Public Heal h, Uni e si y o Copenhagen, Copenhagen, Denma k A icle his o y: Recei ed Oc obe 1, 2019; Accep ed Ma ch 10, 2020 Keywo ds: Sleep pa e ns; Sleep; Adolescen s; In e na ional su eys; Epidemiology; Public Heal h; Age dispa i ies; Socioeconomic di e ences ABSTRACT Pu pose: Insu ficien and poo sleep pa e ns a e common among adolescen s wo ldwide. Up o now, he e idence on adolescen sleep has been mos ly in o med by coun y-specific s udies ha used di e en measu es and age g oups, making di ec compa isons di ficul . C oss-na ional da a on adolescen sleep ha could in o m na ions and in e na ional discussions a e lacking. We examined he sleep pa e ns o adolescen s ac oss 24 coun ies and by gende , age, and a fluence g oups. Me hods: We ob ained sleep da a on 165,793 adolescen s (mean age 13.5 yea s; 50.5% gi ls) in 24 Eu opean and No h Ame ican coun ies om he ecen c oss-sec ional Heal h Beha iou in School-aged Child en su eys (2013e2014 and 2017e2018). Fo each coun y, we calcula ed he age-s anda dized mean in sleep du a ion, iming, and consis ency and he p opo ions mee ing IMPLICATIONS AND CONTRIBUTION The sleep pa e ns o ado- lescen s a y ac oss coun- ies and by sociodemog aphic g oups. Insu ficien sleep on school days is p e alen c oss-na ionally. Na ional policies and s a egies ha Conflic s o in e es : The au ho s ha e no conflic s o in e es o disclose. Disclosu e: This supplemen was suppo ed by he Wo ld Heal h O ganiza ion Eu opean O fice and he Uni e si y o Glasgow. The a icles ha e been pee - e iewed and edi ed by he edi o ial s a o he Jou nal o Adolescen Heal h. The opinions o iews exp essed in his supplemen a e hose o he au ho s and do no necessa ily ep esen he o ficial posi ion o he unde . *Add ess co espondence o: Gene ie e Ga iepy, Ph.D., Depa men o Social and P e en i e Medicine, Uni e si y o Mon eal, 7101 du Pa c, Mon eal, QC H3N 1X9, Canada. E-mail add ess: gene ie [email p o ec ed] (G. Ga iepy). www.jahonline.o g 1054-139X/Ó2020 Published by Else ie Inc. on behal o Socie y o Adolescen Heal h and Medicine. This is an open access a icle unde he CC BY-NC-ND license (h p:// c ea i ecommons.o g/licenses/by-nc-nd/4.0/). h ps://doi.o g/10.1016/j.jadoheal h.2020.03.013 Jou nal o Adolescen Heal h 66 (2020) S81eS88 sleep ecommenda ions on school and nonschool days om sel - epo ed bed imes and wake imes. We conduc ed s a ified analyses by gende , age, and amily a fluence g oup. Resul s: Adolescen sleep pa e ns a ied c oss-na ionally. The a e age sleep du a ion anged be ween 7:47 and 9:07 hou s on school days and be ween 9:31 and 10:22 hou s on nonschool days, and he p opo ion o adolescen s mee ing sleep ecommenda ions anged be ween 32% and 86% on school days and be ween 79% and 92% on nonschool days. Sleep pa e ns by gende and a fluence g oups we e la gely simila , bu olde adolescen s slep less and wen o bed la e on school days han younge adolescen s in all coun ies. Conclusions: The sleep pa e ns o adolescen s a y ac oss coun ies and sociodemog aphic g oups. Insu ficien sleep on school days is common in many coun ies. Public heal h and policy e o s o p omo e heal hy adolescen sleep a e encou aged. Ó2020PublishedbyElse ie Inc.onbehal o Socie y o Adolescen Heal handMedicine.Thisisanopen access a icle unde he CC BY-NC-NDlicense (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/). suppo adolescen sleep a e encou aged. Adolescen sleep is a public heal h conce n in many coun ies. Poo sleep pa e nsdincluding sleep ha is insu ficien , incon- sis en , non es o a i e, dis up ed, o poo ly imed (e.g., la e bed ime)da ec some 30%e70% o adolescen s in Eu ope and No h Ame ica [1,2], whe e a leas one ou h o adolescen s ha e di ficul y alling asleep a nigh and eel i ed du ing he day [1]. Sleep pa e ns o adolescen s di e on school and nonschool days, and hei ci cadian sleepewake cycle is delayed by 2e 3 hou s ela i e o adul s [1,3]. Ea ly school s a imes conflic wi h his ci cadian shi [3,4], leading o unca ed sleep du ing he week [5e7]. On weekends and holidays, adolescen s end o sleep longe and la e [1]. This disc epancy in sleep imes is called social je lag [8] and has been associa ed wi h heal h, beha io al, and cogni i e issues [9e11]. O e all, poo sleep is linked o physical heal h p oblems such as colds, headaches, and obesi y [12,13], emo ional p oblems such as anxie y, dep ession, and suicidal hough s [10,12,14], isky beha io s [15], poo aca- demic pe o mance [9,12], and wo se quali y o li e [13]. Poo sleep also ela es o physical inac i i y, seden a y beha io s including excessi e sc een ime, and poo die [12], unde sco ing he need o include heal hy sleep in o you h heal h p omo ion e o s. The e idence on adolescen sleep is mos ly in o med by coun y-specific s udies ha in ol e di e en measu es and age g oups o e he las ew decades, making di ec compa isons di ficul [1,2,16e18]. None heless, he e a e indica ions ha adolescen sleep a ies ac oss egions and cul u es. A 2011 sys- ema ic e iew o 41 s udies ound la e bed imes in Eu ope han in No h Ame ica (22:46 s. 22:06) and longe sleep on school days (8:26 s. 7:28 hou s) and weekends (10:01 s. 8:48 hou s) [1]. The e iewalso ound ha going o bed abou 2 hou s la e on weekends han on school days was common ac oss egions. Mo eo e , con empo a y in o ma ion on c oss-na ional sleep pa e ns is sca ce and mos ly desc ibes sleep du a ion, whe eas new e idence shows ha o he sleep pa e ns also a ec heal h [11]. Social pa e ns in sleep a e ound be ween age, gende , and socioeconomic g oups. Olde adolescen s go o bed la e , sleep less on school days, and sleep mo e on weekendsda common end in Eu ope, No h Ame ica, and Asia [1,18,19]da ibu able o biological [3,4], beha io al [20], and social changes [21] du ing adolescence. Some s udies ound ha boys sleep less and go o bed la e han gi ls [16,19,22], pa icula ly on weekends [19], whe eas o he s ound ha boys sleep mo e [2,16,17]. Fu he - mo e, s udies om Canada, he U.S., Aus alia, and No way ound ha sho e and poo e quali y sleep is mo e common in lowe socioeconomic g oups [2,16,23], bu compa able da a om o he coun ies a e sca ce. Using da a on sleep collec ed in school-based su eys o ad- olescen s in Eu ope and Canada, we examined c oss-na ional di e ences in sleep du a ion, sleep imes, and social je lag and di e ences by gende , age, and socioeconomic g oups. We hy- po hesized ha sleep pa e ns would a y ac oss coun ies and be ween sociodemog aphic subg oups. Specifically, we an ici- pa ed ha olde adolescen s would go o bed la e , sleep less on school days, and sleep mo e on weekends han younge adoles- cen s, and ha lowe socioeconomic g oups would sleep less han highe socioeconomic g oups, bu we we e agnos ic abou gende di e ences, gi en he mixed findings in he li e a u e. Me hods Sample and p ocedu e The Heal h Beha iou in School-aged Child en (HBSC) s udy is a Wo ld Heal h O ganiza ion collabo a i e c oss-sec ional s udy cu en ly conduc ed in 47 coun ies ac oss Eu ope and No h Ame ica (h p://www.hbsc.o g)[24]. Da a collec ion p ocedu es in all coun ies we e conduc ed in acco dance wi h a s anda d- ized in e na ional p o ocol [24]. Each membe coun y ob ained e hics clea ance o conduc he su ey om a uni e si y-based e iew boa d o equi alen egula o y body. Da a a e collec ed in school se ings e e y 4 yea s om a na ionally ep esen a i e andom clus e sample o 11-,13-, and 15-yea -old adolescen s in each pa icipa ing coun y. The HBSC su ey uses a wo-s age sampling o ec ui s uden s wi hin s a ified samples o schools ha ep esen ed he egional, economic, and public- p i a e dis ibu ion in each coun y. The p ima y sampling uni s a e schools. Mo e de ailed in o ma ion abou me hodology o he HBSC s udy is epo ed elsewhe e [24]. Fo he fi s ime in 2013e2014 and again in 2017e2018, he HBSC su ey included sleep i ems in one o he su ey's op ional ques ionnai e pack- ages. HBSC coun ies could choose o add his op ional package o su ey i ems on sleep o he HBSC manda o y ques ionnai e, depending on hei needs and p io i ies. Twen y- ou coun ies elec ed o include sleep i ems in hei 2013e2014 o 2017e2018 na ional su ey and we e included in his s udy (n ¼185,278). Fo he nine coun ies ha collec ed sleep da a in hei 2013e 2014 and 2017e2018 su eys, we combined he samples o bo h su eys o maximize he use o he a ailable da a and he s a- is ical powe o es ima e subg oup di e ences wi hin each coun y. Sleep pa e ns ac oss he 2013e2014 and 2017e2008 G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88S82 su eys we e compa able. Pa icipan s who did no espond o any sleep i em (n ¼18,489) o did no epo hei age (n ¼996) we e excluded, lea ing a final sample o 165,793 pa icipan s om 24 coun ies (Table 1 and Supplemen a y Table 1). Measu es Sleep pa e ns we e de e mined using ou i ems ha asked, “When do you usually go o bed?”and “When do you usually wake up?”on school days and weekends o holidays (nonschool days). Bed imes anged in hal -hou in e als om “No la e han 21:00” o “2:00 o la e ” o school days and o “4:00 o la e ” o nonschool days. Wake imes anged in hal -hou in e als om “No la e han 5:00” o “8:00 o la e ” o school days and om “No la e han 7:00” o “14:00 o la e ” o nonschool days. We calcula ed sleep du a ion on school and nonschool days using bed and wake imes. Fo each hal -hou in e al, we used he ime a midpoin in ou calcula ions (e.g., 22:15 i he bed ime esponse was “be ween 22:00 and 22:30”). Fo esponses a he ends o he scale, we used he minimum/maximum s a ed ime (e.g., 14:00 i he wake ime esponse was “14:00 o la e ”and 5:00 i he wake ime esponse was “No la e han 5:00”). We calcula ed social je lag using he di e ence in bed imes on nonschool and school days [25]. We also assessed whe he s u- den s me he minimum sleep ecommenda ion o hei age using cu en in e na ional guidelines [26,27], which ecom- mend a leas 9 hou s o hose aged 5e13 yea s (op imal ange 9e11 hou s) and 8 hou s o hose aged 14e17 yea s (op imal ange 8e10 hou s). Socioeconomic s a us (SES) was assessed using he Family A fluence Scale, a checklis o six amily asse s: numbe o ca s, ha ing own bed oom, numbe o compu e s, numbe o ba h- ooms, amily holidays in he pas yea , and ha ing a dishwashe . The Family A fluence Scale is a alida ed scale o compa a i e measu es o amily a fluence o Eu opean and No h Ame ican adolescen s [28]. We ca ego ized s uden s in o e ile g oups o low, medium, o high amily a fluence ela i e o he dis ibu ion o o al sco es om he same age, gende , coun y, and su ey yea . We also used sel - epo ed age and gende (boy o gi l). S a ified analyses used he h ee sampled age g oups in HBSC [24], 11-yea -olds (10.5e12.4 yea s), 13-yea -olds (12.5e 14.4 yea s), and 15-yea -olds (14.5e16.5 yea s). S a is ical analysis All analyses we e un sepa a ely o each coun y. We fi s conduc ed a desc ip i e analysis o he sample cha ac e is ics. We hen es ima ed he age-s anda dized a e age sleep du a- ions, bed imes, wake imes, and social je lag. We s anda dized he esul s o age using he age dis ibu ion o all coun ies combined as a e e ence o accoun o he di e en age dis i- bu ions ac oss coun ies. Finally, we conduc ed s a ified ana- lyses by gende , age, and amily a fluence g oups in each coun y. We s a is ically compa ed di e ences in subg oups using he adjus ed Wald es o con inuous sleep ou comes and he co - ec ed weigh ed Pea son chi-squa e s a is ic o ca ego ical sleep ou comes o accoun o he complex su ey design. Because o he a ying sample sizes ac oss coun ies (n ange om 2,667 o 15,432), we in e p e ed he s a is ical significance wi h cau ion and wi h conside a ion o he public heal h ele ance o he es ima es. Analyses we e done using S a a/SE 15 (S a aCo p LLC, College S a ion, TX). We accoun ed o clus e ing a he school le el and inco po a ed sampling weigh s whe e ele an o ensu e esul s we e ep esen a i e o he coun y using he s y op ion in S a a. Resul s A e age sleep pa e ns on school and nonschool days by coun y a e p esen ed in Figu es 1 and 2, espec i ely, and in Supplemen a y Table 2. School day sleep du a ion anged om 7:47 hou s in Poland o 9:07 hou s in Flemish Belgium, whe eas on nonschool days, i anged om 9:31 hou s in G eece o 10:22 hou s in Flemish Belgium. School day bed imes anged om 21:42 hou s in Flemish Belgium o 23:18 hou s in G eece, whe eas nonschool day bed imes anged om 23:14 hou s in Flemish Belgium o 0:54 hou s in Spain. School day wake imes anged om 06:18 hou s in Hunga y o 07:23 hou s in Spain. Nonschool day wake imes anged om 09:11 hou s in Hunga y o 10:36 hou s in Spain. The pe cen age o adolescen s mee ing sleep ecommenda ions on school days anged om 32.0% in Poland o 86.3% in Flemish Belgium and on nonschool days om 79.2% in G eece o 92.4% in Flemish Belgium (Figu e 3). Social je lag anged om 1:11 hou s in Moldo a o 2:12 hou s in Sweden. Boys epo ed longe sleep du a ion on school days han gi ls in wo hi ds o coun ies (15/24 coun ies), al hough di e ences we e small (10 minu es), e en i s a is ically significan . On nonschool days, gi ls slep longe in all coun ies, om 4 minu es mo e in Swi ze land o 34 minu es mo e in Slo enia. Boys had sligh ly la e bed imes on school days (2e10 minu es) and nonschool days (1e27 minu es). On school days, gi ls woke ea lie han boys in all coun ies, bu on nonschool days, boys woke ea lie han gi ls in abou wo hi ds o coun ies. Al hough boys we e mo e likely o mee sleep ecommenda ions on school days in mos coun ies (17/24 coun ies), gi ls we e mo e likely o mee sleep ecommenda ions on nonschool days in all coun- ies. Boys had abou he same o sligh ly mo e social je lag han Table 1 Su ey yea s and sample size o Heal h Beha iou in School-aged Child en coun ies wi h sleep da a Coun y Su ey yea s Sample size Flemish Belgium 2017e2018 4,305 F ench Belgium 2013e2014, 2017e2018 9,791 Canada 2013e2014 11,295 Czech Republic 2013e2014, 2017e2018 15,432 Denma k 2013e2014, 2017e2018 6,019 Es onia 2013e2014, 2017e2018 8,633 Finland 2013e2014, 2017e2018 7,975 G eece 2017e2018 3,817 Hunga y 2013e2014, 2017e2018 7,397 Iceland 2017e2018 6,967 La ia 2013e2014, 2017e2018 9,887 The Ne he lands 2017e2018 4,634 No way 2017e2018 3,085 Poland 2017e2018 5,205 Po ugal 2017e2018 2,971 Republic o Moldo a 2017e2018 4,682 Sco land 2013e2014, 2017e2018 8,794 Slo akia 2013e2014, 2017e2018 10,736 Slo enia 2013e2014 4,982 Spain 2013e2014 7,576 Sweden 2013e2014 7,511 Swi ze land 2013e2014 2,667 Uk aine 2017e2018 6,600 Wales 2013e2014 4,832 G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88 S83 gi ls, anging om no di e ence in Uk aine o 22 minu es in Sweden. Olde adolescen s slep less in all coun ies (Supplemen a y Table 3). On school days, di e ences be ween he younges and oldes g oups anged om 54 minu es in Slo akia o 1:44 hou s in Slo enia. On nonschool days, he smalles di e ence was 11 minu es in Flemish Belgium, and he g ea es di e ence was 52 minu es in Uk aine. Mos coun ies also saw longe delays in bed imes in olde adolescen s, bo h on school days ( om 52 minu es in Flemish Belgium o 1:35 hou s in Iceland) and nonschool days ( om 1:15 minu es in Flemish Belgium and Uk aine o 2:00 hou s in Sweden). Al hough age di e ences in wake imes on school days a ied conside ably be ween coun- ies, wake imes on nonschool days we e consis en ly la e in olde adolescen s, wi h 15-yea -olds waking 23 minu es la e han 11-yea -olds in Uk aine and up o 1:22 hou s la e in No - way. Age di e ences we e also ound in he pe cen age o ado- lescen s who me sleep ecommenda ions. On school days, 11- yea -olds we e mos likely o ge he ecommended amoun o sleep, anging om 42.8% in Poland o 92.9% in Flemish Belgium, whe eas on nonschool days, 15-yea -olds we e mos likely o mee ecommenda ions, anging om 83.7% in G eece o 96.3% in Iceland. Finally, mo e social je lag was obse ed in he oldes g oup compa ed wi h he younges in all coun ies. Supplemen a y Table 4 shows c oss-na ional sleep pa e ns by amily a fluence g oups. An equal mix o posi i e and nega i e a fluence g adien s was obse ed o sleep du a ion. Howe e , mos di e ences be ween a fluence g oups we e small. On school days, he di e ence in sleep du a ion be ween he leas and mos a fluen adolescen s anged om 31 minu es in Moldo a o þ10 minu es in Sco land and Swi ze land. On nonschool days, sleep du a ion be ween he leas and mos a fluen adolescen s anged om 22 minu es less sleep in Moldo a o þ13 minu es mo e sleep in Sco land. On school days, bed imes we e la e in he high a fluence g oup compa ed wi h he low a fluence g oup in mos coun ies. The g ea es di e - ence was in Moldo a (þ31 minu es). In con as , he high a flu- ence g oup wen o bed ea lie in F ench Belgium, Denma k, Sco land, Wales, and Swi ze land, up o 19 minu es ea lie in Swi ze land. On nonschool days, bed imes we e la e in he high a fluence g oup in hal he coun ies, wi h he la ges di e ence in Moldo a (þ54 minu es) and ea lie in o he coun ies, he la ges di e ence in Sco land (30 minu es). The e was li le a iabili y in wake imes among a fluence g oups. On school days, he di e ence anged om þ5 minu es la e in Canada and Hunga y o 10 minu es ea lie in Swi ze land. The e was also no domina ing end on nonschool days. The di e ence be ween he high and low a fluence g oups anged om þ32 minu es in Moldo a o 18 minu es ea lie in Sco land. On school days, mo e low a fluence adolescen s me sleep ecommenda ions in abou hal he coun ies. The g ea es di e ence was in Moldo a (þ19.1 pe cen age poin s). The high a fluence g oup be e me sleep ecommenda ions in o he coun ies, wi h he highes di e ence ound in Sco land (þ7.2 pe cen age poin s). Simila ends we e obse ed on nonschool days. The p opo ion mee ing sleep ecommenda ions on nonschool days was highe in he high a fluence g oup in some coun ies, wi h he la ges di e ence in F ench Belgium (þ5.2 pe cen age poin s), and highe in he low a fluence g oup in o he coun ies, wi h he la ges di e ence epo ed in Moldo a (þ6.5 pe cen age poin s). Mo e a fluen adolescen s epo ed mo e social je lag in mos coun ies. The g ea es inc ease was in Swi ze land, whe e high a fluence adolescen s had 35 minu es mo e social je lag han low a fluence adolescen s. Con e sely, low a fluence adolescen s had Figu e 1. A e age sleep imes and sleep du a ion on school days by coun y. G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88S84 g ea e social je lag in o he coun ies, up o 25 minu es in F ench Belgium. Discussion This s udy examined sleep du a ion, iming, and consis ency in o e 165,000 adolescen s om 24 p ima ily Eu opean coun- ies. Sleep du a ion anged om 7:47 o 9:07 hou s on school days and om 9:31 o 10:22 hou s on nonschool days. A conside ably smalle p opo ion o s uden s me sleep du a ion ecommenda ions on school days ( ange 32%e86%) han on nonschool days ( ange 79%e92%). Social je lag was obse ed in all coun ies, as bed imes we e consis en ly la e and sleep du- a ions consis en ly longe on nonschool han school days. These di e ences be ween school and nonschool days a e consis en wi h adolescen sleep pa e ns p e iously obse ed in he li e - a u e [1]. In line wi h p e ious s udies, sleep iming and du a ion changed wi h age [1,18,19] bu we e only sligh ly di e en be- ween boys and gi ls [2,16,17,19,22]. Olde adolescen s slep less, wen o bed la e , and expe ienced g ea e social je lag han younge adolescen s. These changes eflec he no ma i e de elopmen al cou se o he ci cadian clock, which shi s owa d e eningness a ound pube y, peaks a a ound age 16 yea s, and shi s back in ea ly adul hood [3,4]. The inc eased demands o schoolwo k and he use o social media and elec onic de ices [20] u he delay sleep imes and add o he biological endency o olde adolescen s o all asleep la e , a a ime when pa en al con ol o e sleep imes diminishes [21]. Howe e , we examined di e ences in sleep pa e ns ac oss age and no pube al de elopmen . Pube al s age was no assessed in mos HBSC coun ies. P e ious s udies no e ha changes in sleep du ing adolescence a e a ibu able o a ious physiological and ho monal changes ela ed o pube y and less so wi h age pe se [29,30]. Fu u e s udies could add nuance o hese analyses by add essing pube al de elopmen in addi ion o age. Al hough we ound li le o e all gende di e ences in sleep pa e ns, he e we e no able coun y-specific di e ences. Fo example, boys slep 10 minu es longe on school days han gi ls in Poland, and gi ls slep 34 minu es longe on nonschool days in Slo enia, sugges ing ha gende di e ences in beha io al pa - e ns, oles, o esponsibili ies may influence adolescen sleep in some coun ies. We obse ed some socioeconomic g adien s in sleep mea- su es, bu hey ended o be small and fluc ua e be ween coun- ies (i.e., posi i e in some coun ies and nega i e in o he s). Ou esul s di e om p e ious s udies epo ing an associa ion be- ween lowe SES and sho e sleep among adolescen s [2,16,23]. Howe e , p e ious e idence s emmed om smalle , na ional samples and used di e en SES measu es. Fu he mo e, a e o- spec i e coho s udy om No way ound ha sleep p oblems we e mo e common in adolescen s who expe ienced downwa d socioeconomic mobili y compa ed wi h s able o upwa d mobili y [31]. The disc epancy wi h ou findings may he e o e poin o me hodological di e ences bu also sugges ha he ela ionship be ween SES and sleep is complex and a ies ac oss na ional and cul u al se ings. This s udy ound significan c oss-na ional a ia ions in sleep pa e ns, sugges ing ha coun y-le el ac o s may play a ole in p o ec ing and p omo ing adolescen sleep. One de e minan may be di e ences in school s a imes. A g owing body o e - idence finds ha la e school s a imes a e associa ed wi h longe sleep du a ions, p ima ily by delaying wake imes [5e7]. Figu e 2. A e age sleep imes and sleep du a ion on nonschool days by coun y. G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88 S85 In e en ion s udies ha e eplica ed hese findings [32], sup- po ing he implemen a ion o his policy o p o ec adolescen sleep. The HBSC su eys did no collec in o ma ion on school s a imes, and na ional da a on school s a imes a e no a ailable. Howe e , a s udy om Canada ound a na ional a e age school s a ime o 08:43 hou s in 2013e2014 [6], whe eas a s udy om he U.S. ound an a e age na ional s a ime o 08:03 hou s in 2011e2012 [7], indica ing a iabili y in school s a imes ac oss coun ies. Commu ing ime o school may also a ec he sleep pa e ns o adolescen s and a y be- ween coun ies [33]. The esul s om a ime use s udy on Ame ican high school s uden s sugges ha each addi ional minu e spen commu ing o school was associa ed wi h 1.3 ewe minu es sleeping [34]. The mode o anspo a ion o commu e o school (e.g., walking s. d i ing) ha is no ma i e o accessible in di e en coun ies may also a ec sleep. In addi ion, ime spen on homewo k is associa ed wi h sho e sleep in adoles- cen s [14] and a ies ac oss coun ies [35]. Mo eo e , cul u al di e ences in bed imes and meal imes [36] and in heal h and li es yle no ms, such as physical ac i i y and heal hy ea ing [20], may con ibu e o di e ences in adolescen sleep pa e ns ac oss coun ies. Finally, geog aphic di e ences could also add o c oss- na ional sleep a ia ions. Fo ins ance, in coun ies si ua ed a highe la i udes, changes in dayligh iming and du a ion h oughou he yea end o co ela e wi h sleep imes, such ha he la e and longe dayligh o he summe ela es o la e bed imes and less sleep compa ed wi h he win e [37]. A no able s eng h o ou s udy is he use o an in e na ional s anda dized p o ocol o da a collec ion, he eby ensu ing high alidi y o c oss-na ional compa isons. By using la ge, na ionally ep esen a i e samples om a o al o 24 coun ies, he p esen s udy adds significan ly o he li e a u e on sociodemog aphic di e ences in adolescen sleeping pa e ns by documen ing a numbe o obus di e ences ac oss na ional se ings and by highligh ing he ele ance o na ional influences o sleep. Ou s udy used sel - epo ed sleep da a. Al hough some s udies ound ha sel - epo ed sleep co ela es highly wi h ac ig aphy da a [38], o he s ha e ound disc epancies ha sug- ges cau ion wi h in e p e a ion [39,40]. Howe e , sleep da a we e measu ed using s anda dized sleep ques ions making ela i e c oss-na ional compa isons mo e likely o be alid despi e imp ecisions. Fu he mo e, sel - epo ed sleep da a gene ally o e es ima e sleep du a ions because sleep onse ends o be la e han bed ime, he e o e esul ing in mo e conse a- i e es ima es in ou s udy. The cu poin s used o define minimal sleep ecommenda ions [26,27] we e based on a sys ema ic e- iew o he bes a ailable e idence [13]. Howe e , much o his e idence was no o high quali y [13], and i is possible ha he cu poin s could change in he u u e as he quan i y and quali y o e idence imp o e. None heless, sleep du a ion ecommen- da ions, e en when based on weake e idence, a e impo an om a popula ion heal h pe spec i e, as hey play an impo an ole in su eillance and public heal h policy o ma ion. Finally, ou da a a e limi ed o s uden s a ending school and may no be gene alizable o adolescen s ou side he school se ing. This s udy p o ides he la ges c oss-na ional accoun o adolescen sleep and sociodemog aphic pa e ning o da e. The esul s show ha , al hough he sleep pa e ns o adolescen s a y ac oss coun ies and sociodemog aphic subg oups, insu fi- cien sleep and poo sleep pa e ns a e p e alen among ado- lescen s in mos coun ies and subg oups. These esul s build on p e ious e idence om coun y-specific s udies and poin o na ional se ings and sociodemog aphic g oups as s ong in- fluences o conside in he p omo ion o heal hy adolescen sleep. Because good sleep is essen ial o se e al aspec s o a heal hy adolescen de elopmen , public heal h policies should Figu e 3. Pe cen age o adolescen s ha mee sleep ecommenda ions on school days and nonschool days by coun y. G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88S86 include c oss-sec o al in e en ions ha a ge adolescen sleep. These policies should consis o modifiable, policy-le el mea- su es ha suppo heal hy adolescen sleep, such as delaying school s a imes and educa ing adolescen s, pa en s, and eache s abou he impo ance o sleep hygiene. Fu he esea ch on c oss-na ional di e ences o o he aspec s o sleep, such as sleeping di ficul ies and subjec i e sleep quali y, is wa an ed, as hey a e impo an componen s o o e all sleep. Fu u e esea ch is also needed o unde s and he c oss-coun y beha io al ac o s and he social, s uc u al, and cul u al de e minan s ha influ- ence sleep pa e ns, such as school and wo k schedules. Acknowledgmen s The Heal h Beha iou in School-aged Child en is an in e na- ional s udy ca ied ou in collabo a ion wi h WHO/EURO. The In e na ional Coo dina o was Candace Cu ie (Uni e si y o S And ews) o he 2013/2014 su ey and Jo Inchley (Uni e si y o Glasgow) o he 2017/2018 su ey. The Da a Bank Manage was P o esso Odd un Samdal (Uni e si y o Be gen). The 2013/2014 and 2017/2018 su eys included in his s udy we e conduc ed by he ollowing p incipal in es iga o s in he 24 coun ies: Flemish Belgium (Ba De Cle cq and Anne Huble ), F ench Belgium (Ka ia Cas e bon and Danielle Pie e), Canada (William Picke and Wendy C aig), Czech Republic (Michal Kalman), Denma k (Me e Rasmussen), Es onia (Leila Oja and Ka in Aas ee), Finland (Jo ma Tynjälä), G eece (Anna Kokke i), Hunga y (Ágnes Ném- e h), Iceland (A saell M. A na sson), La ia (I e a Pudule), he Ne he lands (Gonneke S e ens, Saskia an Do sselae , Wilma Vollebe gh, and Tom de Bog ), No way (Odd un Samdal), Poland (Joanna Mazu ), Po ugal (Ma ga ida Gaspa de Ma os), Republic o Moldo a (Galina Lesco), Sco land (Jo Inchley, Candace Cu ie), Slo akia (And ea Mada aso a Gecko a), Slo enia (Helena Je - icek), Spain (Ca men Mo eno), Sweden (Pe a Lo s ed and Lilly Augus ine), Swi ze land (Ma ina Delg ande-Jo dan, He é Kuendig, and Emmanuelle Kun sche), Uk aine (Olga Balaki e a), and Wales (Ch is Robe s). Funding Sou ces The wo k was suppo ed by he Eu opean Regional De elop- men Fund-P ojec "E ec i e Use o Social Resea ch S udies o P ac ice" (No. CZ.02.1.01/0.0/0.0/16_025/0007294) and by unding om he Technology Agency o he Czech Republic (ÉTA TL01000335) and he Minis y o Educa ion, You h and Spo s, In- e -Excellence, LTT18020 (HBSC Czech Republic); he Public Heal h Agency o Canada (HBSC Canada); he Juho Vainio Founda ion and he Uni e si y o Jy askyla (HBSC Finland); and he Po ugal- Na- ional Founda ion o Science and Technology (HBSC Po ugal). Supplemen a y Da a Supplemen a y da a ela ed o his a icle can be ound a h ps://doi.o g/10.1016/j.jadoheal h.2020.03.013. Re e ences [1] G adisa M, Ga dne G, Dohn H. Recen wo ldwide sleep pa e ns and p oblems du ing adolescence: A e iew and me a-analysis o age, egion, and sleep. Sleep Med 2011;12:110e8. [2] Pa e KA, Qian W, Lea he dale ST. Sleep du a ion ends and ajec o ies among you h in he COMPASS s udy. Sleep Heal h 2017;3:309e16. [3] Randle C, Faßl C, Kalb N. F om la k o owl: De elopmen al changes in mo ningness-e eningness om new-bo ns o ea ly adul hood. Scien ific Rep 2017;7:45874. [4] C owley SJ, Acebo C, Ca skadon MA. Sleep, ci cadian hy hms, and delayed phase in adolescence. Sleep Med 2007;8:602e12. [5] Ma x R, Tanne -Smi h EE, Da ison CM, e al. La e school s a imes o suppo ing he educa ion, heal h, and well-being o high school s uden s. Coch ane Da abase Sys Re 2017;7:CD009467. [6] Ga iepy G, Janssen I, Sen enac M, e al. School s a ime and sleep in Ca- nadian adolescen s. J Sleep Res 2017;26:195e201. [7] Whea on AG, Chapman DP, C o JB. School s a imes, sleep, beha io al, heal h, and academic ou comes: A e iew o he li e a u e. J Sch Heal h 2016;86:363e81. [8] Wi mann M, Dinich J, Me ow M, e al. Social je lag: Misalignmen o biological and social ime. Ch onobiol In 2006;23:497e509. [9] Diaz-Mo ales JF, Esc ibano C. Social je lag, academic achie emen and cogni i e pe o mance: Unde s anding gende /sex di e ences. Ch onobi- ology In 2015;32:822e31. [10] Hende son SEM, B ady EM, Robe son N. Associa ions be ween social je lag and men al heal h in young people: A sys ema ic e iew. Ch onobiology In 2019;36:1316e33. [11] Toui ou Y. Adolescen sleep misalignmen : A ch onic je lag and a ma e o public heal h. J Physiology-Pa is 2013;107:323e6. [12] Shocha T, Cohen-Zion M, Tzischinsky O. Func ional consequences o inadequa e sleep in adolescen s: A sys ema ic e iew. Sleep Med Re 2014; 18:75e87. [13] Chapu J-P, G ay CE, Poi as VJ, e al. Sys ema ic e iew o he ela ionships be ween sleep du a ion and heal h indica o s in school-aged child en and you h. Appl Physiol Nu Me ab 2016;41:S266e82. [14] Becke SP, Langbe g JM, Bya s KC. Ad ancing a biopsychosocial and con ex ual model o sleep in adolescence: A e iew and in oduc ion o he special issue. J You h Adolescence 2015;44:239e70. [15] Sho MA, Webe N. Sleep du a ion and isk- aking in adolescen s: A sys- ema ic e iew and me a-analysis. Sleep Med Re 2018;41:185e96. [16] Keyes KM, Maslowsky J, Hamil on A, e al. The g ea sleep ecession: Changes in sleep du a ion among US adolescen s, 1991-2012. Pedia ics 2015;135:460e8. [17] Hysing M, Pallesen S, S o ma k KM, e al. Sleep pa e ns and insomnia among adolescen s: A popula ion-based s udy. J Sleep Res 2013;22:549e56. [18] Gus a sson ML, Laaksonen C, Salan e a S, e al. Changes in he amoun o sleep and day ime sleepiness: A ollow-up s udy o schoolchild en om ages 10 o 15 yea s. In J Nu s P ac 2019;25:e12689. [19] Olds T, Blunden S, Pe ko J, e al. The ela ionships be ween sex, age, ge- og aphy and ime in bed in adolescen s: A me a-analysis o da a om 23 coun ies. Sleep Med Re 2010;14:371e8. [20] Ba el KA, G adisa M, Williamson P. P o ec i e and isk ac o s o adolescen sleep: A me a-analy ic e iew. Sleep Med Re 2015;21:72e85. [21] Pype E, Ha ing on D, Manson H. Do pa en s’suppo beha iou s p edic whe he o no hei child en ge su ficien sleep? A c oss-sec ional s udy. BMC Public Heal h 2017;17:432. [22] No ell-Cla ke A, Hagquis C. Changes in sleep habi s be ween 1985 and 2013 among child en and adolescen s in Sweden. Scand J Public Heal h 2017;45:869e77. [23] Felden ÉPG, Lei e CR, Rebela o CF, e al. Sleep in adolescen s o di e en so- cioeconomic s a us: A sys ema ic e iew. Re Paul Pedia 2015;33:467e73. [24] Inchley J, Cu ie D, Cosma A, e al. Heal h Beha iou in School-aged Chil- d en (HBSC) s udy p o ocol: Backg ound, me hodology and manda o y i ems o he 2017/18 su ey. S And ews: CAHRU; 2018. [25] Jankowski KS. Social je lag: Sleep-co ec ed o mula. Ch onobiology In 2017;34:531e5. [26] Hi shkowi z M, Whi on K, Albe SM, e al. Na ional Sleep Founda ion’s upda ed sleep du a ion ecommenda ions: Final epo . Sleep Heal h 2015; 1:233e43. [27] T emblay MS, Ca son V, Chapu J-P, e al. Canadian 24-hou mo emen guidelines o child en and you h: An in eg a ion o physical ac i i y, seden a y beha iou , and sleep. Appl Physiol Nu Me ab 2016;41: S311e27. [28] To sheim T, Ca allo F, Le in KA, e al. Psychome ic alida ion o he e ised amily a fluence scale: A la en a iable app oach. Child Indica o s Res 2016;9:771e84. [29] Ca skadon MA, Acebo C. Regula ion o sleepiness in adolescen s: Upda e, insigh s, and specula ion. Sleep 2002;25:606e14. [30] Sadeh A, Dahl RE, Shaha G, e al. Sleep and he ansi ion o adolescence: A longi udinal s udy. Sleep 2009;32:1602e9. [31] Si e sen B, Bøe T, Skogen JC, e al. Mo ing in o po e y du ing childhood is associa ed wi h la e sleep p oblems. Sleep Med 2017;37:54e9. [32] Minges KE, Redeke NS. Delayed school s a imes and adolescen sleep: A sys ema ic e iew o he expe imen al e idence. Sleep Med Re 2016;28: 86e95. [33] Pe ei a EF, Mo eno C, Louzada FM. Inc eased commu ing o school ime educes sleep du a ion in adolescen s. Ch onobiol In 2014;31:87e94. G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88 S87 [34] Voulga is CT, Sma MJ, Taylo BD. Ti ed o commu ing? Rela ionships among jou neys o school, sleep, and exe cise among Ame ican eenage s. J Plann Educa ion Res 2019;39:142e54. [35] O ganisa ion o Economic Co-ope a ion and De elopmen . Does home- wo k pe pe ua e inequi ies in educa ion? Pa is, F ance: OECD Publishing; 2014. [36] Smi h CS, Folka d S, Schmiede RA, e al. In es iga ion o mo ningee ening o ien a ion in six coun ies using he p e e ences scale. Pe sonal Indi idual Di e ences 2002;32:949e68. [37] Ba el K, an Maanen A, Casso J, e al. The sho and long o adolescen sleep: The unique impac o day leng h. Sleep Med 2017;38:31e6. [38] Nascimen o-Fe ei a MV, Collese TS, de Mo aes ACF, e al. Validi y and eliabili y o sleep ime ques ionnai es in child en and adoles- cen s: A sys ema ic e iew and me a-analysis. Sleep Med Re 2016; 30:85e96. [39] Sho MA, G adisa M, Lack LC, e al. The disc epancy be ween ac ig aphic and sleep dia y measu es o sleep in adolescen s. Sleep Med 2012;13:378e 84. [40] Guedes LG, Ab eu GdA, Rod igues DF, e al. Compa ison be ween sel - epo ed sleep du a ion and ac ig aphy among adolescen s: Gende di e ences. Re is a B asilei a de Epidemiologia 2016;19: 339e47. G. Ga iepy e al. / Jou nal o Adolescen Heal h 66 (2020) S81eS88S88