Adolescent alcohol use and parental and adolescent socioeconomic position in six European cities

RESEARCH ARTICLE Open Access Adolescen alcohol use and pa en al and adolescen socioeconomic posi ion in six Eu opean ci ies Ma ina Bosque-P ous 1,2,3,4* , Mi e A. G. Kuipe s 4 , Albe Espel 1,2,5,6 , Ma hias Rich e 7 , A ja Rimpelä 8,9 , Julian Pe elman 10,11 , B uno Fede ico 12 , M. Te esa B ugal 1,2 , Vincen Lo an 13 and An on E. Kuns 4 Abs ac Backg ound: Many isk beha iou s in adolescence a e socially pa e ned. Howe e , i is unclea o wha ex en socioeconomic posi ion (SEP) in luences adolescen d inking in a ious pa s o Eu ope. We examined how alcohol consump ion is associa ed wi h pa en al SEP and adolescen s’own SEP among s uden s aged 14–17 yea s. Me hods: C oss-sec ional da a we e collec ed in he 2013 SILNE s udy. Pa icipan s we e 8705 s uden s aged 14– 17 yea s om 6 Eu opean ci ies. The dependen a iable was weekly binge d inking. Main independen a iables we e pa en al SEP (pa en al educa ion le el and amily a luence) and adolescen s’own SEP (s uden weekly income and academic achie emen ). Mul ile el Poisson eg ession models wi h obus a iance and andom in e cep we e i ed o es ima e he associa ion be ween adolescen d inking and SEP. Resul s: P e alence o weekly binge d inking was 4.2% (95%CI = 3.8–4.6). Weekly binge d inking was no associa ed wi h pa en al educa ion o amily a luence. Howe e , weekly binge d inking was less p e alen in adolescen s wi h high academic achie emen han hose wi h low achie emen (PR = 0.34; 95%CI = 0.14–0.87), and mo e p e alen in adolescen s wi h >€50 weekly income compa ed o hose wi h ≤€5/week (PR = 3.14; 95%CI = 2.23–4.42). These associa ions we e ound o a y acco ding o coun y, bu no acco ding o gende o age g oup. Conclusions: Ac oss he six Eu opean ci ies, adolescen d inking was associa ed wi h adolescen s’own SEP, bu no wi h pa en al SEP. Socio-economic inequali ies in adolescen d inking seem o s em om adolescen s’own si ua ion a he han ha o hei amily. Keywo ds: Alcohol d inking, Adolescence, Socioeconomic ac o s, Eu ope Backg ound Alcohol consump ion anks among he op i e isk ac o s o disease, disabili y and dea h h oughou he wo ld [1]. Adolescence is a pe iod in which many isk beha iou s in- cluding alcohol consump ion a e ini ia ed [2]. P e alence a es among Eu opean adolescen s a y om 6 o 23% o weekly alcohol consump ion [3, 4], and om 27% o 70% o mon hly binge d inking (i.e. d inking la ge amoun s o alcohol in a single occasion) [5]. In adolescen s, equen al- cohol use and binge d inking o en co-occu wi h o he p oblem beha iou s (e.g. academic p oblems, use o o he subs ances, delinquen beha iou , d i ing o iding unde he in luence o alcohol), which may pose signi ican chal- lenges o making a success ul ansi ion om adolescence o adul hood [6–8]. Mo eo e , alcohol use in adolescence p edic s alcohol use and alcoholism in adul hood [9]. Many isk beha iou s in adolescence a e socially pa - e ned. Se e al s udies ha e assessed he possible associ- a ion be ween pa en al socioeconomic posi ion (SEP) and adolescen alcohol consump ion. These s udies yielded in- consis en esul s [10–12]. Acco ding o a e iew o he associa ion be ween SEP and adolescen alcohol consump- ion, mos o he high quali y s udies ound no signi ican associa ion, wo s udies epo ed nega i e associa ions and ou s udies ound posi i e associa ions [10]. Di e ences * Co espondence: [email p o ec ed] 1 Agencia de Salu Pública de Ba celona, Plaça Lesseps 1, 08023 Ba celona, Spain 2 Ins i u d’In es igació Biomèdica San Pau (IIB San Pau), Ca e San An oni Ma ia Cla e 167, 08025 Ba celona, Spain Full lis o au ho in o ma ion is a ailable a he end o he a icle © The Au ho (s). 2017 Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. Bosque-P ous e al. BMC Public Heal h (2017) 17:646 DOI 10.1186/s12889-017-4635-7 in esul s could be due o he use o di e en SEP indica- o s (e.g. pa en al educa ion, pa en al occupa ion o amily income), which may measu e di e en and complemen- a y dimensions o SEP [12]. Inconsis encies be ween s udies migh also be explained by a ia ions in na ional con ex s, including di e ences in pa e ns and meanings o use o alcohol [10, 11]. Some s udies ha e also sug- ges ed ha he ela ionship be ween pa en al SEP and adolescen d inking a ies wi h age and gende [13, 14]. The lack o a consis en associa ion be ween he pa - en al SEP and adolescen alcohol use could be due o adolescence being a pe iod o dec easing pa en al in lu- ence on isk beha iou s [15]. Thus, measu es mo e closely ela ed o he adolescen such as hei academic pe o mance o hei inancial esou ces migh play a mo e impo an ole in adolescen d inking beha iou han socioeconomic cha ac e is ics o he pa en s. P e i- ous s udies showed ha highe adolescen income was associa ed wi h egula alcohol use and consump ion o la ge amoun s [16, 17], and academic achie emen was in e sely associa ed wi h alcohol consump ion [18]. The aim o his s udy was o es ima e he associa ion be ween adolescen d inking and pa en al SEP and ado- lescen s’own SEP among s uden s aged 14–17 yea s, and o analyse which SEP indica o has a s onge asso- cia ion wi h adolescen d inking. We addi ionally ex- plo ed whe he he associa ions a ied acco ding o coun y, gende and age g oup. The pa icula alue o his s udy lies in analysing mul iple SEP indica o s a pa en al and adolescen le el a he same ime, and in using a la ge da ase co e ing ci ies om six di e en Eu opean coun ies. Me hods Design and s udy popula ion The s udy used a c oss-sec ional design wi h da a om he SILNE (Smoking Inequali ies: Lea ning om Na u al Expe imen s) su ey, conduc ed in 2013. The SILNE su - ey was p ima ily aimed a s udying socioeconomic in- equali ies in smoking in di e en Eu opean ci ies [19], bu i also included a iables on alcohol consump ion. The SILNE su ey sample included 50 seconda y schools om six Eu opean ci ies: Namu (Belgium), Tampe e (Finland), Hanno e (Ge many), La ina (I aly), Ame s oo ( he Ne he lands) and Coimb a (Po ugal) (n= 11,015; 79.4% esponse a e). All da a we e sel - epo ed by he s uden s. E hical app o al was ob ained in all coun ies. Me hodo- logical de ails on he SILNE su ey ha e been published elsewhe e [19]. Fo his s udy, we excluded s uden s unde 14 and o e 17 (n= 424) and hose wi h missing in o ma- ion on bo h alcohol measu es (n=12),demog aphic a i- ables (n= 178) and pa en al and adolescen SEP a iables (n= 1696). The inal sample size was o 8705 s uden s. Measu es Dependen a iables The dependen a iable was weekly binge d inking, de- ined as d inking 5 o mo e alcoholic d inks on one oc- casion a leas once a week in he p e ious 12 mon hs (yes/no). Mo eo e , o desc ibe adolescen alcohol con- sump ion, weekly alcohol consump ion (i.e. d inking a leas one alcoholic be e age pe week in he p io 12 mon hs) was also used o pe o m a sensi i i y ana- lysis. In o ma ion on weekly binge d inking and weekly alcohol consump ion was a ailable o 8541 and 8652 in- di iduals, espec i ely. Independen a iables Pa en al SEP We measu ed wo di e en indica o s o pa en al SEP: pa en al educa ion le el and amily a lu- ence, which ha e been p e iously used [4, 10, 12, 20]. Pa en al educa ion le el was measu ed using a coun y- speci ic classi ica ion ha was collapsed in o h ee ca - ego ies: low (p ima y school o lowe le el o seconda y school), middle (comple ed seconda y school o lowe le el college) o high educa ional le el (uni e si y de- g ee). Fo each s uden , he educa ion le el o he pa en wi h he highes a ainmen was used. Family a luence was measu ed wi h he Family A luence Scale (FAS) [21], which was cons uc ed wi h 4 ques ions o he SILNE ques ionnai e: 1) Does you amily own a ca , an o uck?; 2) Do you ha e you own bed oom?; 3) How many compu e s/lap ops/ able s does you amily own?; 4) Du - ing he pas 12 mon hs, how many imes did you a el away on holiday wi h you amily? Ques ion 2 was a ed om 0 o 1 and ques ions 1, 3 and 4 we e a ed om 0 o 2, esul ing in an index wi h alues om 0 o 7. High sco es ep esen ed highe amily a luence. FAS was ca e- go ised due o low numbe s a he ex emes o he scale. Fo easons o s a is ical powe , we used di e en cu -o s o he o al sample analysis han o s a i ied analyses. Adolescen SEP We measu ed wo di e en indica o s o adolescen s SEP, he amoun o money s uden s had a ailable o spend on hemsel es each week and he aca- demic achie emen , which ha e been used in p e ious s udies [22, 23]. S uden weekly income (i.e. he amoun o money he adolescen usually had each week o spend on hemsel es o o sa e om pocke money and jobs) was di ided in o i e ca ego ies: €0–5, €6–10, €11–20, €21–50 o >€50. Due o low numbe s in some ca ego ies in he s a i ied analysis, we used a a iable wi h only h ee g oups: €0–5, €6–20 and >€20. Adolescen aca- demic achie emen was based on he s uden ’s ma ks du ing he p e ious school yea , (classi ied in i e ca - ego ies: insu icien , low, a e age, good o high achie e- men . Fo he analysis s a i ying by coun y, we used a Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 2 o 10 a iable wi h only h ee ca ego ies: low (insu icien and low ma ks), a e age and good (good and high ma ks). O he independen a iables We measu ed con- ounde s likely o p edic adolescen d inking, such as age (in yea s), gende (male s. emale) and mig an backg ound [24]. We dis inguished be ween na i e back- g ound, mixed backg ound (one pa en bo n in ano he coun y han he coun y o esidence), and mig an backg ound (bo h pa en s bo n in a o eign coun y). In he s a i ied analysis age was classi ied in o wo g oups: younge s uden s (14–15 yea s old) and olde s uden s (16–17 yea s old). S a is ical analysis Fi s , we desc ibed he dis ibu ion o he independen a iables in he sample and es ima ed he age-adjus ed p e alence o weekly binge d inking and weekly alcohol consump ion o all subg oups. We also es ed o poly- cho ic co ela ions be ween he dependen a iables, as he a iables we e ca ego ical [25]. As we ound ha weekly binge d inking and alcohol consump ion we e highly co ela ed ( = 0.88), we decided o ocus he s udy on weekly binge d inking and conduc a sensi i i y analysis wi h weekly alcohol consump ion as he dependen a iable. We calcula ed polycho ic co ela ions be ween he di - e en adolescen and pa en al SEP a iables o assess po en ial mul icollinea i y in he eg ession model. We ound ha he co ela ions be ween hese a iables we e weak (be ween −0.10 and +0.21). To es ima e he associ- a ions be ween weekly binge d inking and pa en al edu- ca ion, amily a luence, s uden weekly income and academic achie emen , we i mul ile el Poisson eg es- sion models wi h obus a iance and andom in e cep (2 le els: s uden and school). These models yielded P e alence Ra ios (PR) and hei 95% con idence in e - als (95%CI). We ied o i a h ee-le el model includ- ing he ci y as he highes le el, bu he model i poo ly due o he ac ha school a iabili y explained he di - e ences be ween coun ies and also because he e we e ew coun ies in he s udy. We used Poisson eg ession models wi h obus a iance ins ead o logis ic eg es- sion models as hey yielded PR, which ha e he ad an- age o e Odds Ra ios o ha e a conc e e in e p e a ion e en when p e alence a es a e high [26]. We i se e al models in a ious s eps. In each s ep, we adjus ed o gende , age and mig an backg ound. In s ep 1 we es ima ed he a iance among schools. In s ep 2 we i se e al models including in each model only one SEP measu e (pa en al educa ion, amily a luence, s uden weekly income o academic achie emen ). In s ep 3 we in- cluded all SEP measu es in one model simul aneously. We conduc ed a sensi i i y analysis using mul iple impu a ion in addi ion o he comple e-case analysis app oach in o de o assess whe he he me hod used o deal wi h missing da a in luenced he esul s. To es whe he he associa ions we e consis en among gende , age and coun y, we also i he same e- g ession models pe gende , age and coun y. All ana- lyses we e conduc ed using STATA 13. Resul s Table 1 p esen s he dis ibu ion o he s udy popula ion and he p e alence o weekly binge d inking and weekly alcohol consump ion o each o he independen a i- ables. The mean age o pa icipan s was 15.2 yea s and 53.5% we e gi ls. O e all, age-adjus ed p e alence o weekly binge d inking and alcohol consump ion we e 4.2% (95%CI = 3.8–4.6) and 11.3% (95%CI = 10.7–12.0), espec i ely, wi h ele an a ia ions be ween coun ies. The p e alence o weekly binge d inking a ied om 1.1% (95%CI = 0.7–1.9) in Finland o 6.1% (95%CI = 5.1–7.3) in Belgium, and he p e alence o weekly alcohol consump- ion a ied om 2.1% (95%CI = 1.4–3.1) in Finland o 19.8% (95%CI = 18.1–21.7) in I aly. We did no obse e consis en di e ences in he p e alence o alcohol con- sump ion be ween ca ego ies o pa en al SEP indica o s, whe eas p e alence a es we e clea ly highe in adoles- cen s wi h lowe academic achie emen and highe weekly income. The age-adjus ed p e alence o weekly binge d inking in adolescen s who ecei ed €0–5/week and >€50/week was 2.4% and 10.7%, espec i ely. In adoles- cen s whose academic achie emen was he lowes and he highes , he age-adjus ed p e alence o weekly binge d inking was 7.0% and 1.5%, espec i ely. The esul s o he mul ile el Poisson eg ession con- i m he indings o he desc ip i e analysis (Table 2). When adjus ed by he con ol a iables (age, gende , mi- g an backg ound and school) and o o he SEP a i- ables, we did no ind an associa ion be ween pa en al SEP indica o s and adolescen alcohol consump ion, bu we ound a dose- esponse ype associa ion be ween weekly binge d inking and adolescen SEP. Adolescen s wi h high academic achie emen we e less likely o binge d ink weekly han hose wi h insu icien academic achie emen (PR = 0.34; 95%CI = 0.14–0.87). The asso- cia ion ound o s uden weekly income was in he op- posi e di ec ion: adolescen s wi h a high weekly income (>€50/week) had a highe isk o weekly binge d inking han hose wi h a low weekly income (≤€5/week) (PR = 3.14; 95%CI = 2.23–4.42). I is impo an o high- ligh ha he PR o adolescen SEP did no change in he ully adjus ed model, which means ha pa en al SEP did no ac as a con ounde in he ela ionship be ween adolescen SEP and alcohol use. The esul s s a i ied by gende and age g oup a e p e- sen ed in Addi ional ile 1: Table S1 and Addi ional ile 2: Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 3 o 10 Table S2, espec i ely. We did no ind signi ican di e - ences be ween g oups, al hough he associa ions seemed s onge in gi ls and in younge s uden s. Table 3 shows he esul s o he analyses s a i ied by coun y. In gene al, coun y-speci ic esul s we e simila o hose o he whole sample. Howe e , we ound some Table 1 Dis ibu ion o he indi idual independen a iables and age-adjus ed p e alence o weekly binge d inking and weekly alcohol consump ion (i.e. d inking a leas one alcoholic be e age pe week) among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013 All pa icipan s Weekly binge d inking Weekly alcohol consump ion N % Age-adjus ed p e alence 95%CI Age-adjus ed p e alence 95%CI Gende Female 4658 53.5 2.9 (2.4–3.5) 8.4 (7.7–9.3) Male 4047 46.5 5.6 (4.9–6.3) 14.6 (13.5–15.7) Ci y (coun y) Namu (Belgium) 1653 19.0 6.1 (5.1–7.3) 15.7 (14.1–17.4) Tampe e (Finland) 1016 11.7 1.1 (0.7–1.9) 2.1 (1.4–3.1) Hanno e (Ge many) 1060 12.2 4.9 (3.6–6.7) 7.6 (6.0–9.6) La ina (I aly) 1874 21.5 5.7 (4.7–6.8) 19.8 (18.1–21.7) Ame s oo (Ne he lands) 1548 17.8 4.6 (3.5–5.9) 13.8 (12.0–15.7) Coimb a (Po ugal) 1554 17.8 1.9 (1.4–2.6) 4.7 (3.9–5.8) Mig an backg ound Na i e 6988 80.3 4.3 (3.8–4.8) 12.1 (11.4–12.9) Mixed couples 913 10.5 4.8 (3.6–6.5) 10.8 (8.9–12.9) Bo h pa en s immig an s 804 9.2 3.1 (2.1–4.5) 6.4 (4.9–8.2) Pa en al educa ion le el Low le el 1177 13.5 5.3 (4.1–6.8) 11.5 (9.7–13.6) Middle le el 3497 40.2 4.7 (4.0–5.5) 12.7 (11.5–13.9) High le el 4031 46.3 3.7 (3.1–4.4) 10.8 (9.8–11.9) Family A luence Scale 0–2 837 9.6 5.5 (4.0–7.4) 11.4 (9.3–13.7) 3 1360 15.6 4.1 (3.1–5.2) 9.7 (8.3–11.4) 4 2253 25.9 3.7 (3.0–4.5) 9.3 (8.2–10.5) 5 1522 17.5 3.8 (2.9–4.9) 12.0 (10.5–13.8) 6–7 2733 31.4 4.5 (3.7–5.4) 13.7 (12.4–15.1) Academic achie emen Insu icien (<50%) 313 3.6 7.0 (4.9–10.0) 14.2 (11.0–18.1) Low (50–59%) 1117 12.8 5.5 (4.3–7.0) 14.0 (12.1–16.2) A e age (60–69%) 3616 41.5 4.9 (4.2–5.6) 12.8 (11.8–14.0) Good (70–84%) 2784 32.0 3.2 (2.6–3.9) 9.8 (8.7–10.9) High (>85%) 875 10.1 1.5 (0.9–2.6) 5.3 (4.0–6.9) S uden weekly income 0–5€2063 23.7 2.4 (1.8–3.1) 6.9 (5.9–8.0) 6–10 €2084 23.9 2.0 (1.5–2.7) 7.3 (6.3–8.4) 11–20 €2070 23.8 3.8 (3.1–4.7) 11.3 (10.0–12.7) 21–50 €1623 18.7 5.9 (4.8–7.2) 15.5 (13.8–17.3) >50€865 9.9 10.7 (8.7–13.1) 22.8 (20.1–25.9) To al 8705 100.0 8541 8652 To al p e alence 4.2 (3.8–4.6) 11.3 (10.7–12.0) Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 4 o 10 di e ences be ween coun ies. Pa en al SEP measu es we e no associa ed wi h adolescen d inking in i e coun ies. I alian s uden s om highe a luen amilies we e a lowe isk o weekly binge d inking (PR = 0.69; 95%CI = 0.54–0.88). In addi ion, academic achie emen and s uden weekly income we e associa ed wi h binge d inking in 2 and 4 ou o 6 ci ies s udied, espec i ely. In he Ge man ci y, he associa ion wi h academic achie emen seemed o be in he opposi e di ec ion o he o he i e ci ies, bu i was no signi ican . The asso- cia ion be ween s uden weekly income and binge d ink- ing in he Du ch and he Finish ci ies was no s a is ically signi ican , bu he poin es ima es we e in he same di ec ion o he o he coun ies. Finally, in a sensi i i y analysis, we epea ed hese ana- lyses using weekly alcohol consump ion as he dependen a iable. The esul s o he mul ile el Poisson eg ession models o weekly alcohol consump ion a e shown in Addi ional ile 3: Table S3. The s a i ied analyses a e shown in Addi ional ile 4: Table S4, Addi ional ile 5: Table S5 and Addi ional ile 6: Table S6. The es ima es o bo h alcohol measu es we e simila in all analyses. When conduc ing a sensi i i y analysis wi h mul iple impu a ion o missing alues, we ob ained e y simila associa ions. Discussion The main esul s o his s udy we e: 1) Pa en al SEP mea- su es (pa en al educa ion and amily a luence) we e no associa ed wi h adolescen d inking; 2) High s uden weekly income and low academic achie emen we e asso- cia ed wi h highe likelihood o alcohol consump ion; 3) The same associa ions we e obse ed o di e en d ink- ing measu es (weekly binge d inking and weekly alcohol consump ion), and o di e en sexes and age g oups; and 4) Al hough he ela ionship be ween SEP indica o s and binge d inking a ied be ween coun ies, in mos coun- ies adolescen SEP was associa ed wi h alcohol con- sump ion, while pa en al SEP was no . Table 2 P e alence a ios (PR) o binge d inking es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013 S ep 2 S ep 3 PR 95%CI PR 95%CI Pa en al educa ion le el Low le el 1 1 Middle le el 0.96 (0.75–1.23) 0.93 (0.73–1.19) High le el 0.95 (0.64–1.42) 0.93 (0.64–1.35) Family A luence Scale (FAS) 0–21 1 3 0.93 (0.60–1.44) 0.93 (0.59–1.46) 4 0.95 (0.63–1.41) 0.89 (0.59–1.34) 5 0.96 (0.64–1.43) 0.89 (0.58–1.39) 6–7 1.09 (0.74–1.61) 0.98 (0.66–1.46) Academic achie emen Insu icien (<50%) 1 1 Low (50–59%) 0.75 (0.42–1.33) 0.73 (0.41–1.31) A e age (60–69%) 0.67 (0.36–1.27) 0.66 (0.36–1.21) Good (70–84%) 0.50 (0.28–0.88) 0.49 (0.29–0.83) High (>85%) 0.34 (0.13–0.91) 0.34 (0.14–0.87) S uden weekly income 0–5€11 6–10 €0.93 (0.64–1.38) 0.95 (0.65–1.38) 11–20 €1.59 (1.11–2.29) 1.55 (1.07–2.24) 21–50 €2.14 (1.48–3.11) 2.12 (1.47–3.05) >50€3.10 (2.23–4.30) 3.14 (2.23–4.42) Va iabili y (% change in a iabili y) a 0.458 (10.3) S ep 2 included weekly binge d inking, one SEP indica o and was adjus ed by age, gende and mig an backg ound in le el 1 and school in le el 2. S ep 3 also included all SEP indica o s a Va iabili y o he emp y model (s ep 1), which included only weekly binge d inking was 0.511. % change in a iabili y was calcula ed using he ollowing o mula: [( a iabili y s ep 1 - a iabili y cu en s ep)/( a iabili y s ep 1)]×100 Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 5 o 10 Table 3 P e alence o binge d inking and p e alence a ios o he socioeconomic posi ion (SEP) a iables by coun y, es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013 Namu (Belgium) Tampe e (Finland) Hanno e (Ge many) La ina (I aly) Ame s oo (Ne he lands) Coimb a (Po ugal) % PR 95%CI % PR 95%CI % PR 95%CI % PR 95%CI % PR 95%CI % PR 95%CI Pa en al educa ion le el Low le el 6.6 1 no obse a ions 2.8 1 7.4 1 5.9 1 4.1 1 Middle le el 6.5 0.87 (0.71–1.07) 1.8 1 5.4 2.11 (0.40–11.15) 5.1 0.83 (0.61–1.13) 4.2 0.74 (0.32–1.71) 3.4 0.83 (0.52–1.32) High le el 7.7 1.26 (0.71–2.25) 0.6 0.70 (0.21–2.36) 3.1 1.47 (0.33–6.59) 4.3 0.85 (0.53–1.35) 2.9 0.63 (0.26–1.54) 2.0 0.51 (0.17–1.50) Family A luence Scale (FAS) 0–4 6.3 1 1.3 1 4.6 1 6.3 1 3.5 1 2.7 1 5–7 8.3 1.18 (0.92–1.52) 1.1 0.95 (0.27–3.37) 3.0 0.75 (0.44–1.25) 4.7 0.69 (0.54–0.88) 3.7 1.20 (0.77–1.86) 3.2 1.31 (0.77–2.22) Academic achie emen Low (<60%) 9.2 1 4.2 1 3.3 1 14.2 1 4.3 1 6.5 1 A e age (60–69%) 7.9 0.92 (0.44–1.91) 0.3 0.06 (0.01–0.54) 4.9 1.62 (0.67–3.94) 7.9 0.72 (0.39–1.30) 4.1 1.02 (0.66–1.56) 2.4 0.47 (0.17–1.26) Good (>70%) 5.2 0.68 (0.45–1.02) no obse a ions 2.9 1.55 (0.66–3.66) 2.6 0.30 (0.11–0.78) 2.8 0.93 (0.61–1.40) 1.8 0.49 (0.18–1.35) S uden weekly income 0–5€4.1 1 no obse a ions 1.3 1 3.3 1 2.3 1 2.1 1 6–20 €5.4 1.29 (0.90–1.85) 0.9 1 1.9 1.34 (0.30–6.04) 5.4 1.51 (0.92–2.50) 1.2 0.43 (0.13–1.48) 2.7 1.31 (0.43–3.96) >20€12.1 2.51 (1.49–4.24) 2.2 1.94 (0.94–4.01) 6.4 3.23 (1.00–10.49) 10.5 2.50 (1.28–4.87) 6.2 1.35 (0.63–2.90) 6.9 2.89 (1.29–6.47) Each model included weekly binge d inking, all SEP indica o s and was adjus ed by age, gende and mig an backg ound in le el 1 and school in le el 2 %p e alence; PR p e alence a io; 95%CI 95% con idence in e al Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 6 o 10 E alua ion o po en ial limi a ions Fi s , alcohol consump ion measu es we e es ima ed based on sel - epo ed da a. As a esul , p e alence a es o weekly binge d inking and alcohol consump ion may be unde es ima ed due o social desi abili y bias, as indi- iduals d inking highe amoun s o alcohol end o unde es ima e hei consump ion [27]. Besides, i is also possible ha some adolescen s may exagge a e hei al- cohol consump ion o gain social app o al among pee s, bu his bias may be educed by he use o a sel - comple ed anonymised ques ionnai e. Howe e , o a i- bu e ou key esul s o such biases, one would ha e o assume ha hey a e un ela ed o pa en al SEP bu s ongly ela ed o own academic achie emen and in- come. This seems unlikely. Mo eo e , as alcohol use was epo ed o e lec consump ion in he p e ious yea , he equency o binge d inking and alcohol consump ion may also be unde es ima ed due o ecall bias. Second, pa en al SEP was epo ed by s uden s ins ead o pa en s hemsel es. As a consequence, we could no ask in de ail abou he amily income o pa en al occu- pa ion s a us. Mo eo e , pa en al educa ion le el was he only a iable wi h a ele an p opo ion o missing da a (a ound 12%). This may in oduce bias, as non- esponse a es end o be highe among people wi h low SEP [28]. Howe e , we conduc ed a sensi i i y analysis based on mul iple impu a ion o missing alues, and we ound ha he esul s we e no sensi i e o he way in which missing alues we e deal wi h. Finally, he ac ha all SEP measu es we e sel - epo ed may also be a sou ce o social desi abili y bias, as child en may epo highe le els o SEP o gain social app o al. P e alence o adolescen alcohol use We ound subs an ial di e ences in he p e alence o al- cohol consump ion be ween coun ies, wi h he highes age-adjus ed p e alence o weekly binge d inking in Belgium (6.1%) and he lowes in Finland (1.1%). Al- hough he lowe p e alence o alcohol use in Finish ad- olescen s may be explained by a s eep decline in alcohol use and in oxica ion in adolescen s a e 2011 [29], ac- o s ela ed o he cul u al con ex o alcohol use and he di e en d inking mo i es in each coun y [30], as well as alcohol con ol policies and hei en o cemen , may explain he ele an di e ences in p e alence be- ween coun ies. In his sense, Finish alcohol policies a e among he s ic es in Eu ope [31]. The p e alence o alcohol use in all coun ies is lowe han he one es ima ed in he HBSC su ey o 2009– 2010 [32]. In I aly o example, we ound ha he p e a- lence o weekly alcohol use was 13% (95%CI = 11–14) and 29% (95%CI = 26–33) in gi ls and boys, espec i ely, whe eas in he HBSC su ey he p e alence was 26% (95%CI = 23–29) and 39% (95%CI = 36–42), espec i ely. A possible explana ion o his lowe p e alence may be ha bo h s udies co e di e en pe iods o ime and ha ou da a co e ed single ci ies while he HBSC su ey used ep esen a i e na ional samples, which included u al a eas whe e d inking a es end o be highe han in u ban a eas [7]. Pa en al SEP and adolescen d inking Ou indings sugges ha , in gene al, he e is no associ- a ion be ween pa en al SEP and adolescen d inking, which is in line wi h a e iew ha ound no clea pa e n in his associa ion [10]. Howe e , o he s udies ha e ound an associa ion be ween pa en al SEP and adoles- cen d inking in some o hese coun ies [33–35]. When we s a i ied by coun y, weekly binge d inking was nega i ely associa ed wi h FAS in I aly, whe eas weekly adolescen d inking was posi i ely associa ed wi h pa en- al educa ion in Belgium and wi h FAS in he Ne he lands. These esul s a e pa ly in line wi h esul s on he HBSC su ey, which ound no di e ences in weekly alcohol use by FAS in Belgium, Finland and I aly, bu ound a posi i e associa ion in he Ne he lands [20]. Besides, his lack o associa ion be ween adolescen d inking and pa en al SEP in he o e all sample did no change depending on he alcohol d inking measu e. The esul s om a s udy in Slo ak s uden s we e in ag ee- men wi h ou s: no associa ion was ound be ween pa - en al educa ion and bo h equency o alcohol use and d unkenness [36]. Adolescen SEP and adolescen d inking Wi h espec o he s uden weekly income, we ound a s ong posi i e associa ion: he mo e a ailable money, he highe he isk o binge d ank weekly. This inding is in line wi h p e ious s udies [22, 37]. In English school- child en and US college s uden s, ha ing mo e money o spend was associa ed wi h an inc eased likelihood o d inking equen ly and o binge d inking [22, 37]. This s ong ela ionship implies ha alcohol is no only a - ec ed by access o social sou ces, bu ha he abili y o buy om comme cial sou ces is impo an as well [38]. Adolescen s wi h g ea e access o money o spend may be able o pu chase subs ances mo e easily [10, 39]. Howe e , he obse ed associa ion may also be due o e e se causali y as adolescen s who d ink mo e o en need mo e money o buy alcohol, hey may look o ways o ob ain mo e money (e.g. wo king o asking o mo e pocke money) [40]. P obably bo h p ocesses ein o ce each o he : he mo e you d ink, he mo e money you need; and he mo e money you ha e, he mo e you d ink. Longi udinal s udies ha ollow young people h oughou hei adolescence a e needed o as- sess which o hese p ocesses con ibu e mos o he alcohol-income associa ion. Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 7 o 10 High academic achie emen appea s o be a p o ec - i e ac o o alcohol consump ion. S uden s wi h highe achie emen may be mo e o ien ed owa d he u u e and, he e o e, be mo e likely o p o ec hei heal h. P e ious s udies ound ha ha ing high aca- demic goals, such as planning o g adua e om uni- e si y, is a p o ec i e ac o o subs ance use among adolescen s [41]. Fu he mo e, adolescen s wi h highe ma ks a e pe haps less likely o engage in a e school social ac i i ies, as hey may spend mo e ime s udy- ing and ha e, consequen ly, less social in ol emen . Mo eo e , nega i e school expe iences (including low academic achie emen and low mo i a ion) a e known isk ac o s o subs ance use [23, 42]. Ano he pos- sible explana ion could be ha subs ance use may also con ibu e o academic di icul ies and hus a ec academic pe o mance [18, 43, 44]. Due o he c oss- sec ional na u e o ou da a, we could no sepa a e social causa ion explana ions (SEP a ec ing alcohol use) om social selec ion explana ions (alcohol use a ec ing SEP). Ou esul s sugges he need o u he esea ch using longi udinal da a o iden i y he causal mechanism d i - ing he associa ions be ween adolescen d inking and s uden weekly income and academic achie emen , e- spec i ely. Besides, esea ch on socioeconomic inequal- i ies in adolescen d inking should include measu es o adolescen s’own SEP as hey appea o be much mo e p edic i e o adolescen alcohol use han pa en al SEP indica o s. A diminished alcohol a ailabili y and accessi- bili y in adolescen s should be ein o ced h ough s ic- e alcohol policies, such as inc easing alcohol axes o highe con ol o alcohol selling, among o he e ec i e policies [45–47]. Conclusions Low academic achie emen and high s uden weekly in- come a e bo h s ongly associa ed wi h highe adoles- cen d inking. Pa en al educa ional le el and amily a luence a e no consis en ly associa ed wi h adolescen alcohol consump ion. This sugges s ha socioeconomic inequali ies in adolescen d inking may o igina e om adolescen s’own si ua ion a he han ha o hei am- ily. These esul s imply ha heal h p og ammes o p e- en adolescen d inking in schools should ocus especially on child en wi h school p oblems and low academic pe o me s, jus as i has been sugges ed o obacco [48]. The impo an ole o s uden weekly in- come s esses he impo ance o g ea e a en ion, bo h in public heal h esea ch and p ac ice, on how pa en s could in luence hei child en’s d inking pa e ns and o he heal h- ela ed beha iou h ough p o ision o pocke money. Addi ional iles Addi ional ile 1: Table S1. P e alence a ios (PR) o weekly binge d inking by gende es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013. (DOC 66 kb) Addi ional ile 2: Table S2. P e alence a ios (PR) o weekly binge d inking by g oup o age es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013. (DOC 56 kb) Addi ional ile 3: Table S3. P e alence a ios (PR) o d inking a leas one alcoholic be e age pe week es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013. (DOCX 14 kb) Addi ional ile 4: Table S4. P e alence a ios (PR) o d inking a leas one alcoholic be e age pe week by gende es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14–17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013. (DOCX 15 kb) Addi ional ile 5: Table S5. P e alence a ios (PR) o d inking a leas one alcoholic be e age pe week by g oup o age es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14– 17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013. (DOCX 15 kb) Addi ional ile 6: Table S6. P e alence o d inking a leas one alcoholic be e age pe week and p e alence a ios o he socioeconomic posi ion (SEP) a iables by coun y, es ima ed wi h mul ile el Poisson eg ession models wi h obus a iance among 14– 17 yea s-old s uden s om 6 Eu opean ci ies pa icipa ing in he SILNE su ey, 2013. (DOCX 16 kb) Acknowledgemen s No applicable. Funding This s udy is pa o he SILNE-R p ojec , which is suppo ed by he Eu opean Union’s Ho izon 2020 esea ch and inno a ion p og amme, unde g an ag eemen 635056; and o he p ojec ‘Tackling socio-economic inequali ies in smoking (SILNE)’, which is unded by he Eu opean Commission, Di ec o a e-Gene al o Resea ch and Inno a ion, unde he FP7-Heal h-2011 p og amme, wi h g an ag eemen numbe 278273. This wo k was also sup- po ed by he Spanish Ne wo k on Addic i e Diso de s [g an numbe RD12/ 0028/0018]. The s ay o Ma ina Bosque-P ous a he Uni e si y o Ams e dam was pa ially unded by an Exchange Awa d om he Eu opean Founda ion o Alcohol Resea ch. The unding sou ces had no in ol emen in he s udy design, he collec ion, analysis and in e p e a ion o da a he w i ing o he epo and he decision o submi he a icle o publica ion. This pape is pa o he doc o al disse a ion o Ma ina Bosque-P ous, a he Uni e si a Pompeu Fab a. A ailabili y o da a and ma e ials The da ase s used and/o analysed du ing he cu en s udy a ailable om he co esponding au ho on easonable eques . Au ho s’con ibu ions MBP, MK, AE and AK concep ualized and designed he s udy. MBP ca ied ou he analyses and d a ed he ini ial manusc ip . All au ho s ou side o Spain con ibu ed o iginal da a om he SILNE su ey. MK, AE, MR, AR, JP, BF, TB, VL and AK e iewed he manusc ip , and all au ho s app o ed he inal manusc ip as submi ed. E hics app o al and consen o pa icipa e E hical app o al was ob ained in all coun ies. The e hics commi ees ha app o ed he p ojec in each coun y we e: he Azienda Uni à Sani a ia Locale F osinone (I aly) (App o al’s e e ence numbe : 862, app o ed on 13/ 11/2012), he Medical E hical Commi ee o he AMC (Ne he lands) (App o al’s e e ence numbe : W12_256#12.17.0290), he E hics Commi ee Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 8 o 10 o he Tampe e egion (Finland) (Fa ou able S a emen e e ence numbe : 10/2012), he E hics commi ee o Ma in-Lu he -Uni e si y Halle-Wi enbe g (Ge many) (App o al’s e e ence numbe : 2012–112, app o ed on 13/12/ 2012), he commission d’É hique Biomédicale (Belgium) (App o al’s e e ence numbe : 2012/09OCT/461) and he Gene al Di ec o a e o Educa ion (Po ugal) (App o al’s e e ence numbe : Re numbe 0338600001, app o ed on 02/11/2012). Consen o pa icipa e was ob ained om pa icipan s and hei pa en s in all ci ies (mo e de ailed da a on e e ence [19]). Da a on he SILNE p ojec a e no openly a ailable, bu he owne s o he SILNE da a se a e ep esen ed in he eam o au ho s and M Bosque-P ous ecei ed hei pe mission o use he da a o he pu poses o his pape . Consen o publica ion No applicable. Compe ing in e es s The au ho s decla e ha hey ha e no compe ing in e es s. Publishe ’sNo e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions. Au ho de ails 1 Agencia de Salu Pública de Ba celona, Plaça Lesseps 1, 08023 Ba celona, Spain. 2 Ins i u d’In es igació Biomèdica San Pau (IIB San Pau), Ca e San An oni Ma ia Cla e 167, 08025 Ba celona, Spain. 3 Uni e si a Pompeu Fab a, Ca e Doc o Aiguade 80, 08003 Ba celona, Spain. 4 Depa men o Public Heal h, Academic Medical Cen e , Uni e si y o Ams e dam, 1100 DD Ams e dam, Ams e dam, he Ne he lands. 5 Cen os de In es igación Biomédica en Red. Epidemiología y Salud Pública (CIBERESP), Calle Melcho Fe nández Almag o 3-5, 28029 Mad id, Spain. 6 Depa amen de Psicobiologia i Me odologia en Ciències de la Salu , Facul a de Psicologia, Uni e si a Au ònoma de Ba celona, Campus UAB Bella e a, 08193 Ce danyola del Vallès, Spain. 7 Ins i u e o Medical Sociology (IMS), Medical Facul y, Ma in Lu he Uni e si y Halle-Wi enbe g, Magdebu ge S . 8, 06112 Halle (Saale), Ge many. 8 School o Heal h Sciences and PERLA - Tampe e Cen e o Childhood, You h and Family Resea ch, Uni e si y o Tampe e, 33014 Tampe e, Finland. 9 Depa men o Adolescen Psychia y, Tampe e Uni e si y Hospi al, -33521 Tampe e, FI, Finland. 10 Escola Nacional de Saúde Pública, Uni e sidade NOVA de Lisboa, 1600-560 Lisbon, Po ugal. 11 Cen o de In es igação em Saúde Pública, 1600-560 Lisbon, Po ugal. 12 Depa men o Human Sciences, Socie y and Heal h, Uni e si y o Cassino and Sou he n Lazio, 03043 Cassino, I aly. 13 Ins i u e o Heal h and Socie y, Uni e si é Ca holique de Lou ain, 3016-1200 B ussels, Belgium. Recei ed: 23 No embe 2016 Accep ed: 26 July 2017 Re e ences 1. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adai -Rohani H, e al. A compa a i e isk assessmen o bu den o disease and inju y a ibu able o 67 isk ac o s and isk ac o clus e s in 21 egions, 1990-2010: a sys ema ic analysis o he global bu den o disease s udy 2010. Lance . 2012;380:2224–60. 2. Jackson KM, She KJ, Coope ML, Wood PK. Adolescen alcohol and obacco use: onse , pe sis ence and ajec o ies o use ac oss wo samples. Addic ion. 2002;97:517–31. 3. G een MJ, Leyland AH, Swee ing H, Benze al M. Socioeconomic posi ion and adolescen ajec o ies in smoking, d inking, and psychia ic dis ess. J Adolesc Heal h. 2013;53:202–208.e2. 4. Rich e M, Kun sche E, de Looze M, P ö ne T-K. T ends in socioeconomic inequali ies in adolescen alcohol use in Ge many be ween 1994 and 2006. In J Public Heal h. 2013;58:777–84. 5. Danielsson A-K, Wennbe g P, Hibell B, Romelsjö A. Alcohol use, hea y episodic d inking and subsequen p oblems among adolescen s in 23 Eu opean coun ies: does he p e en ion pa adox apply? Addic ion. 2012; 107:71–80. 6. Ellickson PL, Tucke JS, Klein DJ. Ten-yea p ospec i e s udy o public heal h p oblems associa ed wi h ea ly d inking. Pedia ics. 2003;111:949–55. 7. Fon -Ribe a L, Ga cia-Con inen e X, Pé ez A, To es R, Sala N, Espel A, e al. D i ing unde he in luence o alcohol o d ugs among adolescen s: he ole o u ban and u al en i onmen s. Accid Anal P e . 2013;60:1–4. 8. Schulenbe g JE, Maggs JL. A de elopmen al pe spec i e on alcohol use and hea y d inking du ing adolescence and he ansi ion o young adul hood. J S ud Alcohol Suppl. 2002:54–70. 9. Pi känen T, Lyy a A-L, Pulkkinen L. Age o onse o d inking and he use o alcohol in adul hood: a ollow-up s udy om age 8-42 o emales and males. Addic ion. 2005;100:652–61. 10. Hanson MD, Chen E. Socioeconomic s a us and heal h beha io s in adolescence: a e iew o he li e a u e. J Beha Med. 2007;30:263–85. 11. Lems a M, Benne NR, Neudo C, Kuns A, Nannapaneni U, Wa en LM, e al. A me a-analysis o ma ijuana and alcohol use by socio-economic s a us in adolescen s aged 10-15 yea s. Can J Public Heal h. 2008;99:172–7. 12. Rich e M, Ve eecken CA, Boyce W, Maes L, Gabhainn SN, Cu ie CE. Pa en al occupa ion, amily a luence and adolescen heal h beha iou in 28 coun ies. In J Public Heal h. 2009;54:203–12. 13. Bachman JG, O’Malley PM, Johns on LD, Schulenbe g JE, Wallace JM. Racial/ e hnic di e ences in he ela ionship be ween pa en al educa ion and subs ance use among U.S. 8 h-, 10 h-, and 12 h-g ade s uden s: indings om he moni o ing he u u e p ojec . J S ud Alcohol D ugs. 2011;72:279–85. 14. Salonna F, an Dijk JP, Gecko a AM, Slesko a M, G oo ho JW, Reijne eld SA. Social inequali ies in changes in heal h- ela ed beha iou among Slo ak adolescen s aged be ween 15 and 19: a longi udinal s udy. BMC Public Heal h. 2008;8:57. 15. Wes P, Swee ing H. E idence on equalisa ion in heal h in you h om he wes o Sco land. Soc Sci Med. 2004;59:13–27. 16. Jensen R, Lle as-Muney A. Does s aying in school (and no wo king) p e en een smoking and d inking? J Heal h Econ. 2012;31:644–57. 17. Jones SC, Magee CA. The ole o amily, iends and pee s in Aus alian adolescen ’s alcohol consump ion. D ug Alcohol Re . 2014;33:304–13. 18. Kou a K, Papado assilaki K, Kalpou zaki P, Ka ga zi M, Roumelio aki T, Koukouli S. Adolescen d inking, academic achie emen and leisu e ime use by seconda y educa ion s uden s in a u al a ea o C e e. Heal h Soc Ca e Communi y. 2012;20:61–9. 19. Lo an V, So o VE, Al es J, Fede ico B, Kinnunen J, Kuipe s M, e al. Smoking in school-aged adolescen s: design o a social ne wo k su ey in six Eu opean coun ies. BMC Res No es. 2015;8:91. 20. Cu ie C, Gabhainn SN, Godeau E, Robe s C, Smi h R, Cu ie D, e al. Inequali ies in young people’s heal h: HBSC in e na ional epo om he 2005/2006 su ey. Wo ld Heal h O ganiza ion Copenhagen; 2008. 21. Cu ie C, Molcho M, Boyce W, Hols ein B, To sheim T, Rich e M. Resea ching heal h inequali ies in adolescen s: he de elopmen o he heal h beha iou in school-aged child en (HBSC) amily a luence scale. Soc Sci Med. 2008;66:1429–36. 22. Bellis MA, Hughes K, Mo leo M, Tocque K, Hughes S, Allen T, e al. P edic o s o isky alcohol consump ion in schoolchild en and hei implica ions o p e en ing alcohol- ela ed ha m. Subs Abuse T ea P e Policy. 2007;2:15. 23. B yan AL, Schulenbe g JE, O’Malley PM, Bachman JG, Johns on LD. How academic achie emen , a i udes, and beha io s ela e o he cou se o subs ance use du ing adolescence: a 6-yea . Mul iwa e Na ional Longi udinal S udy J Res Adolesc. 2003;13:361–97. 24. Sa asa-Renedo A, So do L, Pulido J, Gui a A, González-González R, Hoyos J, e al. E ec o immig a ion backg ound and coun y-o -o igin con ex ual ac o s on adolescen subs ance use in Spain. D ug Alcohol Depend. 2015; 153:124–34. 25. Ong AD. Van Dulmen MH. Ox o d handbook o me hods in posi i e psychology: Ox o d Uni e si y P ess New Yo k; 2007. 26. Ba os AJD, Hi aka a VN. Al e na i es o logis ic eg ession in c oss-sec ional s udies: an empi ical compa ison o models ha di ec ly es ima e he p e alence a io. BMC Med Res Me hodol. 2003;3:21. 27. K umpal I. De e minan s o social desi abili y bias in sensi i e su eys: a li e a u e e iew. Qual Quan . 2013;47:2025–47. 28. Ekholm O, Gundgaa d J, Rasmussen NKR, Hansen EH. The e ec o heal h, socio-economic posi ion, and mode o da a collec ion on non- esponse in heal h in e iew su eys. Scand J Public Heal h. 2010;38:699–706. 29. Kinnunen J, Pe e L, Lind o s P. Hanna Ollila, A ja Rimpelä. The adolescen heal h and li es yle su ey. Tobacco and subs ance use among Finnish adolescen s in 1977–2015. Sosiaali- ja e eysminis e iön: Helsinki; 2015. 30. Kun sche E, Gabhainn SN, Robe s C, Windlin B, Vieno A, Bend sen P, e al. D inking mo i es and links o alcohol use in 13 Eu opean coun ies. J S ud Alcohol D ugs. 2014;75:428–37. Bosque-P ous e al. BMC Public Heal h (2017) 17:646 Page 9 o 10