Full Te ms & Condi ions o access and use can be ound a
h ps://www. and online.com/ac ion/jou nalIn o ma ion?jou nalCode= ady20
In e na ional Jou nal o Adolescence and You h
ISSN: 0267-3843 (P in ) 2164-4527 (Online) Jou nal homepage: h ps://www. and online.com/loi/ ady20
Academic pe o mance and adolescen smoking in
6 Eu opean ci ies: he ole o iendship ies
Pie e-Oli ie Robe , Mi e A. G. Kuipe s, Ka ha ina Ra hmann, I ene Moo ,
Jaana M. Kinnunen, A ja Rimpelä, Julian Pe elman, B uno Fede ico, Ma hias
Rich e , An on E. Kuns & Vincen Lo an
To ci e his a icle: Pie e-Oli ie Robe , Mi e A. G. Kuipe s, Ka ha ina Ra hmann, I ene Moo ,
Jaana M. Kinnunen, A ja Rimpelä, Julian Pe elman, B uno Fede ico, Ma hias Rich e , An on E.
Kuns & Vincen Lo an (2019) Academic pe o mance and adolescen smoking in 6 Eu opean
ci ies: he ole o iendship ies, In e na ional Jou nal o Adolescence and You h, 24:1, 125-135,
DOI: 10.1080/02673843.2018.1475288
To link o his a icle: h ps://doi.o g/10.1080/02673843.2018.1475288
© 2018 The Au ho (s). Published by In o ma
UK Limi ed, ading as Taylo & F ancis
G oup.
View supplemen a y ma e ial
Published online: 28 May 2018. Submi you a icle o his jou nal
A icle iews: 1238 View C ossma k da a
h ps://doi.o g/10.1080/02673843.2018.1475288
Academic pe o mance and adolescen smoking in 6 Eu opean
ci ies: he ole o iendship ies
Pie e-Oli ie Robe a, Mi e A. G.Kuipe sb, Ka ha inaRa hmannc,d, I eneMoo c,
Jaana M.Kinnunene, A jaRimpeläe, , JulianPe elmang, B unoFede icoh,
Ma hiasRich e c, An on E.Kuns b and Vincen Lo an a
aIns i u e o heal h and socie y (I ss), uni e si é ca holique de lou ain, B ussels, Belgium; bdepa men o Public
heal h, academic Medical cen e , uni e si y o ams e dam, ams e dam, he ne he lands; cMedical acul y, Ins i u e o
Medical sociology (IMs), Ma in lu he uni e si y halle-Wi enbe g, halle (saale), Ge many; d acul y o ehabili a ion
sciences, echnical uni e si y do mund, do mund, Ge many; eschool o heal h sciences, uni e si y o ampe e,
ampe e, inland; depa men o adolescen Psychia y, Pi käniemi hospi al, ampe e uni e si y hospi al, ampe e,
inland; gescola nacional de saúde Pública and cen o de In es igação em saúde Pública, uni e sidade noVa de
lisboa, lisbon, Po ugal; hdepa men o human sciences, socie y and heal h, uni e si y o cassino and sou he n
lazio, cassino, I aly
ABSTRACT
Poo academic pe o mance is a known isk ac o o adolescen smoking,
ye he associa ion emains unclea , as he ole o social ies has been a ely
examined. Ou s udy aims o in es iga e he ole o iendship ies in his
associa ion. In a sample o 11,015 adolescen s, aged 14 o17, in 50 schools
wi hin six Eu opean ci ies (SILNE-su ey, 2013), we used mul ile el models
o analyse he media ing e ec o he composi ion o iendship ies and
school ypes on he associa ion be ween academic pe o mance and
smoking. Resul s show smoking was mo e p e alen in adolescen s wi h
lowe academic pe o mance han wi h highe . This associa ion was s onge
in non- oca ional schools han in oca ional. Adolescen s ended o ha e
iendship ies wi h someone sha ing he same smoking s a us and academic
pe o mance. Finally, iendship ne wo ks a e pa e ned bo h on smoking
and academic pe o mance. This sugges s he educa ional en i onmen
con ibu es o u u e socio-economic inequali ies in smoking among young
people.
In oduc ion
Among adolescen s, poo academic pe o mance is associa ed wi h a g ea e isk o smoking ini ia ion,
mo e equen smoking, a highe numbe o ciga e es smoked, and ewe a emp s o qui smoking
(Go, Tucke , G een, Polla d, & Kennedy, 2012; Hu, Lin, & Keele , 1998; Ka p, O'loughlin, Pa adis, Hanley, &
Di anza, 2005; Kinnunen e al., 2016; Kun z & Lampe , 2013; Me cken, Snijde s, S eglich, & de V ies, 2009).
High smoking a es among poo academic pe o ming adolescen s ha e been obse ed in c oss-sec-
ional s udies and in longi udinal s udies o smoking onse , and ac oss high income coun ies (B adley
& G eene, 2013; Hu e al., 1998; Ka p e al., 2005; Kinnunen e al., 2016; Me cken, Snijde s, S eglich, &
de V ies, 2009; Schnoh , K eine , Rasmussen, Due, & Dide ichsen, 2009). Academic pe o mance in pa
ARTICLE HISTORY
ecei ed 11 ap il 2018
accep ed8 May 2018
KEYWORDS
daily smoking; academic
pe o mance; adolescen s;
schools; iendship ies
© 2018 he au ho (s). Published by In o ma uK limi ed, ading as aylo & ancis G oup.
his is an open access a icle dis ibu ed unde he e ms o he c ea i e commons a ibu ion 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 he o iginal wo k is p ope ly ci ed.
CONTACT Pie e-oli ie obe pie e-oli ie [email p o ec ed]
he supplemen al da a o his a icle is a ailable online a h ps://doi.o g/10.1080/02673843.2018.1475288.
OPEN ACCESS
In e na Ional Jou nal o adolescence and You h
2019, Vol. 24, No. 1, 125–135
is s ongly ela ed o socio-economic oppo uni ies la e in li e (Lo an e al., 2016). The p e en ion o
smoking in poo ly pe o ming adolescen s is he e o e c ucial in o de o dec ease smoking inequali ies
in adul s. Howe e , he e is a lack o unde s anding why adolescen s who pe o m poo ly in school a e
so much mo e likely o s a smoking.
Some p e ious s udies ha e sugges ed ha poo academic pe o me s a e mo e o en smoke s
due o con ounding ac o s, such as low amily socio-economic backg ound (Ali & Dwye , 2009) and
a highe p e alence o smoking among amily membe s (Go e al., 2012; Me cken, Snijde s,S eglich,
Va iainen, &de V ies, 2009; Simons-Mo on, Ha os, & Haynie, 2004). O he s udies ha e sugges ed
ha he associa ion be ween academic pe o mance and smoking esul s om ma ginaliza ion o poo
pe o ming s uden s. Adolescen s who pe o m less well may eel ma ginalized by eache s and by hei
be e -pe o ming pee s (Els ad, 2010). They a e less likely o ecei e social, emo ional, and academic
suppo om o he s. This makes hem mo e likely o clus e oge he and engage in de ian beha iou ,
such as iola ion o school ules and smoking, as a way o c ea e a social iden i y (Chen & Hsiao, 2009).
Clus e ing o poo pe o ming s uden s may o a la ge pa be exp essed in iendship ies, bu he
ole o such ies has no been ex ensi ely s udied in he exis ing li e a u e (Lo an e al., 2015, 2016).
Adolescen s no only end o c ea e and main ain iendship ies wi h o he adolescen s a a simila
le el o academic pe o mance (Huang, So o, Fujimo o, & Valen e, 2014; Kobus, 2003; Lomi, Snijde s,
S eglich, & To ló, 2011; Me cken, Snijde s, S eglich, & de V ies, 2009; Schae e , Haas, & Bishop, 2012; Wool ,
Po s, Pa el, & MCManus, 2012), bu also wi h o he s wi h a simila smoking s a us (Me cken, Snijde s,
S eglich, & de V ies, 2009, Go e al., 2012; Schae e e al., 2012; Seo & Huang, 2012; DeLay, Lau sen,
Kiu u, Salmela-A o, & Nu mi, 2013). Based on Dimaggio’s heo y o social ne wo k-induced inequali y,
adolescen s’ social ies o s uden s wi h simila smoking beha iou and simila le el o pe o mance
con ibu e o he associa ion be ween academic pe o mance and smoking. This, in u n, could inc ease
ea ly socio-economic smoking inequali ies (DiMaggio & Ga ip, 2012).
This kind o clus e ing may also be a ec ed by school ype, as some schools ack s uden s solely
acco ding o hei academic pe o mance ( an de We ho s , 2011; Ra hmann e al., 2016). School ypes,
school op ions (e.g. spo , sciences, a s, e c.), and class oom assignmen s se e as a p ima y il e o
pee in luence because hey acili a e physical and social p oximi y among adolescen s who sha e
simila cha ac e is ics, including smoking (Go e al., 2012; Me cken e al., 2009; Seo & Huang, 2012) and
academic pe o mance (Huang e al., 2014; Kobus, 2003; Lomi e al., 2011; Me cken, Snijde s, S eglich,
& de V ies, 2009; Schae e e al., 2012; Wool e al., 2012). School ype, which is imposed by he educa-
ional sys em, in luences he composi ion o iendship ies among s uden s. Adolescen s wi h highe
academic pe o mance a e mo e likely o be a end schools ha ha e a g ea e academic ocus and a e
mo e o ien ed owa ds uni e si y deg ees, while adolescen s wi h lowe academic pe o mance may
a end schools wi h a oca ional o ien a ion (Huisman, an de We ho s , & Monshouwe , 2012; Van de
We ho s & Mijs, 2010). T acking o s uden s acco ding o hei academic pe o mance may p omo e
inequali ies no only wi h ega d o academic pe o mance and aspi a ions (Buchmann & Dal on, 2002)
bu also o smoking (Doku, Koi usil a, Rainio, & Rimpelä, 2010; Ra hmann e al., 2016). Rich e (2007)
showed ha s uden s in oca ional-o ien a ion schools ha e a ou imes highe isk o smoking han
s uden s who a e en olled in schools wi h he highes (uni e si y) educa ional o ien a ion (Rich e &
Leppin, 2007). We he e o e expec ed ha he link be ween smoking and low academic pe o mance
would a y acco ding o school ype.
The aim o his s udy was o es o wha ex en he associa ion be ween academic pe o mance
and adolescen smoking is explained by he pa e ning o adolescen s’ social ela ionships based on
smoking and academic pe o mance. We add essed h ee speci ic ques ions:
(1) Wha is he associa ion be ween academic pe o mance and smoking among adolescen s in
Eu ope?
(2) Does he composi ion o social ies explain he associa ion be ween academic pe o mance
and smoking?
126 P.-O. ROBERT ET AL.
(3) Does he associa ion be ween academic pe o mance and smoking a y acco ding o school
ype?
Me hods
This s udy used da a om he 2013 ‘SILNE’ su ey (‘Tackling Smoking Inequali ies: Lea ning om Na u al
Expe imen s’). SILNE is a school-based social ne wo k su ey o adolescen s in wo g ades co espond-
ing o 14–17yea -olds and was conduc ed in six Eu opean ci ies (Namu , Belgium; Tampe e, Finland;
Hano e , Ge many; La ina, I aly; Ame s oo , Ne he lands; and Coimb a, Po ugal). In each ci y an a e -
age o 8 schools and 1800 adolescen s we e ec ui ed. A o al o 163 schools we e con ac ed and 50
ag eed o pa icipa e. A o al o 13,870 adolescen s egis e ed in hese schools we e in i ed o ill in
a w i en ques ionnai e abou hei social ela ionships in school, heal h beha iou s, and sociodemo-
g aphic cha ac e is ics. A o al o 11,015 adolescen s (pa icipa ion a e o 79.4%) pa icipa ed, yielding
57,094 collec ed ies be ween hose adolescen s. In each ci y, e hical app o al was ob ained om
na ional o local o ganiza ions. A de ailed desc ip ion o he design and concep s o he SILNE s udy
has been published elsewhe e (Lo an e al., 2015).
Measu es
Daily smoking was de ined as epo ing ha ing smoked a leas 1 ciga e e pe day in he las 30days.
All o he esponden s we e conside ed non(-daily) smoke s.
Adolescen academic pe o mance was based on he s uden ’s sel - epo ed ma ks du ing he p e i-
ous yea : ‘Which o he ollowing bes desc ibes you ma ks du ing he pas yea ?’. This a iable included
i e ini ial alues based on each coun y’s academic pe o mance assessmen sys em. These alues we e
g ouped in o h ee e ile ca ego ies: low, a e age and high.
Adolescen s we e asked o nomina e up o 5 iends (also e e ed o ‘al e s’) wi hin he wo g ades.
Responden s answe ed he ques ion, ‘who a e you bes and closes iends?’ (Lo an e al., 2015). To
iden i y hei pee s, adolescen s we e handed a s uden di ec o y which con ained he names o all
s uden s en olled in he wo g ades, alphabe ically o each class and g ade. One code was assigned o
each name and esponden s we e asked o use he codes. In Finland, he names o adolescen s had o be
w i en on he ques ionnai e and he esea che s coded hem a e wa ds (name-gene a o app oach).
Rela ionships es ablished be ween adolescen s in ela ion o smoking beha iou and academic pe -
o mance we e desc ibed a he indi idual le el: (1) he p opo ion o iendship ies o low academic
achie e s o he o al numbe o iendship nomina ions ( he numbe o ou going ies ha a s uden has)
and (2) he p opo ion o iendship ies o daily smoke s o he o al numbe o iendship nomina ions.
In he main analysis, he indices we e compu ed o ecip ocal iendship ies. In a sensi i i y analysis
non- ecip ocal iendship ies we e also included. Recip ocal iendships a e gene ally conside ed as
in ol ing mo e us and social capi al han non- ecip ocal iendships (Ri e a, Sode s om, & Uzzi, 2010).
To cap u e he composi ion o ins i u ional social ies imposed by he educa ional sys em, we ca e-
go ized he s uden s in oca ional s. non- oca ional school ypes. The non- oca ional school ype is
de ined as he uni e si y-p epa a o y ack and he oca ional school ype is de ined as ha o s uden s
who en e he labou ma ke a e seconda y school and be o e he age o 20. As Finland does no ha e
acking in he included age- ange, we c ea ed a ‘no- acking’ ca ego y o ha coun y. Supplemen a y
Table 1 desc ibes he alloca ion o he di e en acks pe school ype and pe coun y (Huisman e al.,
2012; Moo e al., 2014).
Co a ia es included he numbe o smoke s among amily membe s in he household, socio-eco-
nomic s a us, age, and sex. Socio-economic s a us was de ined by he pa en al educa ional le el.
Pa en al educa ion was classi ied in o low, middle, high, and unknown, acco ding o he educa ion
sys em o each coun y (Lo an e al., 2015).
INTERNATIONAL JOURNAL OF ADOLESCENCE AND YOUTH 127
Da a analysis
A e excluding 1,139 obse a ions due o missing da a, we we e le wi h 9,876 ull eco ds.
We i s abula ed smoking s a us and academic pe o mance by socio-demog aphic g oups. We
hen desc ibed he composi ion o iendship ies acco ding o he adolescen ’s own smoking s a us
and academic pe o mance le el. We used mul ile el logis ic eg ession models ( andom in e cep ) o
quan i y he associa ion be ween academic pe o mance and smoking. Media ion o he associa ion
by he composi ion o iendship ies was es ed. In Model 1, we eg essed he academic pe o mance
a iable on daily smoking. In Model 2, we added 3 indices o he composi ion o iendship ies: he
composi ion o ins i u ional social ies imposed by he educa ional sys em (school ypes), iendship
ies wi h daily smoke s, and iendship ies wi h low-pe o ming pee s. In Model 3, we con olled Model
2 o gende , age, pa en s’ educa ion, and he numbe o smoke s among amily membe s. S a is ical
analyses we e ca ied ou wi h SAS 9.3. As pa icipa ion di e ed sligh ly be ween schools, we weigh ed
he da a by he in e se o he pa icipa ion a e.
Resul s
Table 1 desc ibes he sample and he dis ibu ion o daily smoking and o academic pe o mance.
Olde adolescen s, s uden s a ending oca ional schools, and s uden s om lowe socio-economic
backg ounds we e mo e likely o be smoke s and o ha e low academic pe o mance. Daily smoking
Table 1.sample cha ac e is ics and dis ibu ion o daily smoking and academic pe o mance, sIlne, 2013, o e all pe cen age and
chi-squa e es (n=9876).
a“no acking”= inland.
Co a ia es % NDaily smoke (%) χ2P alue
Low academic
pe o me (%) χ2P alue
Gende        
Male 49.7 5071 14.3 10.3 0.001 20.6 41.5 <.001
emale 50.3 5132 12.1   15.7  
Age g oup        
12–14 27.1 2770 5.7 417.3 <.001 12.8 205.8 <.001
15 38.3 3903 10.3   15.1  
16+ 34.6 3530 22.3   25.7  
Ci ies        
namu (Be) 17.8 1818 19.2 166.0 <.001 23.3 298.1 0.004
hano e (de) 20.2 2064 9.5   18.8  
ame s oo (nl) 25.4 2594 9.0   10.7  
la ina (I ) 7.3 740 21.1   7.8  
coimb a (P ) 16.8 1709 15.4   19.3  
ampe e ( I) 12.5 1278 11.7   29.1  
School ypes        
non- oca ional 56.1 5668 7.7 407.9 <.001 14.2 166.8 <.001
Voca ional 31.2 3153 23.2   20.6  
‘no acking’a12.7 1278 11.7   29.1  
Mo he ’s educa ion        
low 15.9 1625 17.0 67.2 <.001 22.4 120.3 <.001
Medium 32.6 3328 15.0   19.9  
high 32.8 3343 9.7   12.4  
o he /unknown 18.7 1906 12.9   21.4  
Fa he ’s educa ion        
low 18.6 1902 17.5 84.5 <.001 23.3 151.0 <.001
Medium 28.6 2919 14.6   19.1  
high 31.1 3176 9.1   11.7  
o he /unknown 21.6 2207 13.6   21.7  
Academic pe o mance        
low pe o mance 18.7 1849 23.1 271.0 <.001   
a e age pe o mance 43.2 4264 13.6     
high pe o mance 38.1 3763 7.2     
To al   13.2%   18.7%  
128 P.-O. ROBERT ET AL.

was mo e equen among s uden s wi h lowe academic pe o mance han among s uden s wi h high
academic pe o mance (23.1 s. 7.2%, χ2=271, p<0.001).
Table 2 desc ibes he composi ion o adolescen s’ iendships ies acco ding o hei own academic
pe o mance and smoking s a us. Each cell is he a e age pe cen age o ecip ocal ies o smoke s
(o o low pe o me s) as a p opo ion o he o al numbe o ecip ocal ies. Adolescen s had mo e
iendship ies o adolescen s wi h a simila smoking s a us and wi h a simila academic pe o mance.
The pe cen age o smoking al e s was highe among daily smoke s ( om 40 o 54%) han among
non(-daily) smoke s ( om 5 o 11%), independen o hei academic pe o mance. This associa ion was
s onge among high pe o me s ( a io=8.1) han among low pe o me s ( a io=4.71). I also a ied
acco ding o he academic pe o mance. The pe cen age o daily-smoking iends was highe among
lowe pe o me s han among highe pe o me s ( a io=1.34).
Table 2 also shows ha he pe cen age o ies o daily smoke s inc eased as academic pe o mance
dec eased, bo h among smoke s and non-smoke s. A simila pa e n was obse ed o iendship ies
o s uden s wi h low academic pe o mance. Among lowe pe o me s, be ween 45% and 31% o
ecip ocal iends we e low pe o me s, as agains 16 o 9% among high pe o me s.
Supplemen a y Table 2 shows a eplica ion o Table 2 wi h all ies ( ecip ocal and non- ecip ocal).
The same pa e n was ound, bu wi h less p onounced di e ences, sugges ing ha he s eng h o
ies os e s he associa ion be ween ego-smoking and al e s’ smoking s a us.
Table 3 p esen s he associa ion be ween daily smoking and academic pe o mance by school ypes.
In all coun ies he e was a dose- esponse associa ion be ween academic pe o mance and daily smok-
ing: he lowe he academic pe o mance, he highe he p e alence o daily smoking. S uden s in
oca ional school ype ha e a highe p e alence o daily smoking han hose in non- oca ional school
ypes. O e all, he isk o smoking among low-pe o ming adolescen s compa ed o high pe o me s
was highe o adolescen s in ‘no- acking’ schools and he non- oca ional ype o school han o he
oca ional ype o school.
Table 4 p esen s he mul ile el logis ic eg ession models. In Model 1, low academic pe o mance
(OR 3.86, 95% CI 3.28–4.55) was associa ed wi h daily smoking. In Model 2, shows ha he highe he
pe cen age o iendship ies o daily smoke s, he highe he odds o daily smoking (OR=16.53, 95%
CI: 13.36–20.45). A highe pe cen age o iendship ies wi h low-pe o ming pee s was also associa ed
wi h highe odds o daily smoking (OR=1.89, 95% CI: 1.50–2.39). The associa ion be ween low compa ed
o high academic pe o mance and daily smoking dec eased om 3.86 (Model 1) o 2.26 (Model 2). This
means ha a pa o he associa ion be ween academic pe o mance and daily smoking was a ibu ed
by he composi ion o iendship ies. Mo eo e , adolescen s en olled in oca ional schools compa ed
wi h hose in non- oca ional schools we e mo e likely o smoke (OR=2.28, 95% CI: 1.87–2.76). Model
3 included con ounde s sex, age, pa en al educa ion, and he numbe o amily membe s who a e
smoke s. The isk o daily smoking o low pe o me s dec eased sligh ly be ween Model 2 and Model
3 ( om OR 2.26 o OR 1.99), which sugges s ha amily backg ound and adolescen cha ac e is ics
Table 2. iendship ecip ocal ies wi h smoke s and wi h low academic pe o me s by smoking s a us and academic pe o mance
o ego: a e age pe cen age o he numbe o ies and a io, sIlne in e na ional su ey 2013 (n=11,021).
* a io o iendship ies wi h daily smoke s among daily smoke s / iendship ies wi h daily smoke s among non-daily-smoke s.
  F iendship al e s in % o o al No. o ies
Smoking s a us o ego Composi ion o iendship ies
Low
academic
pe o mance
A e age
academic
pe o mance
High
academic
pe o mance
Ra io
low/
high
daily smoke s iendship ies wi h daily smoke s 53.7% 43.0% 39.9% 1.34
 iendship ies wi h low pe o me s  45.3% 24.6% 15.8% 2.86
non-daily-
smoke s
iendship ies wi h daily smoke s 11.4% 7.9% 4.9% 2.32
  iendship ies wi h low pe o me s 31.0% 16.1% 9.1% 3.40
a io*  4.71 5.44 8.14 
INTERNATIONAL JOURNAL OF ADOLESCENCE AND YOUTH 129
sligh ly con ound he associa ion be ween adolescen s’ own pe o mance and daily smoking. F iendship
ies and school ype emained s ongly ela ed o smoking, independen o amily backg ound and
adolescen cha ac e is ics.
Discussion
Main indings
The o igin o he associa ion be ween smoking and poo academic pe o mance in adolescen s has
emained unclea so a . Using an o iginal in e na ional social-ne wo k s udy ac oss six coun ies, we
in es iga ed he ole o social ies in his associa ion, looking a bo h he indi idual le el and he dyad
le el. We ound ha iendship ies we e homophilous bo h in ela ion o smoking and in ela ion o
academic pe o mance. Adolescen s who a e daily smoke s ha e mo e iends who a e smoke s han
non-smoke s and low-pe o ming adolescen s ha e mo e ies o o he low-pe o ming adolescen s.
Adolescen s en olled in oca ional schools a e mo e likely o smoke han hei coun e pa s in non- o-
ca ional o non- acking schools. This composi ion o iendship ies and school ype pa ly explains
he associa ion be ween smoking and academic pe o mance. Al hough smoking is mo e p e alen
in oca ional schools han in non- oca ional schools, low academic pe o mance is mo e s ongly
Table 3.associa ion o daily smoking wi h academic pe o mance pe school ype: pe cen age and o , sIlne in e na ional s udy
o adolescen heal h, 2013.
ao =odds a io o low pe o me s s. high pe o me s, adjus ed o age and sex wi h 95% con idence In e al (cI). all he P- alues
a e unde 0.001.
b‘no- acking’= ampe e ( inland).
 Daily smoke s (%) pe school ype
Academic pe o mance ‘No- acking’bNon- oca ional Voca ional
low pe o mance 25.6% 16.5% 29.4%
a e age pe o mance 7.8% 8.2% 23.4%
high pe o mance 1.6% 3.9% 18.0%
To al 11.1% 7.5% 23.1%
OR (95%CI)a21.90 (9.74–49,23) 3.25 (2.43–4.36) 1.72 (1.35–2.20)
Table 4. isk o daily smoking acco ding o academic pe o mance, school ype, and composi ion o iendship ies. odds a io om
he logis ic eg ession, sIlne in e na ional su ey o adolescen s, 2013.
no es: Model 1 eg essed he academic pe o mance a iable on daily smoking.
Model 2 added 3 indices o iendship- ies composi ion and andom in e cep a he ci y le el.
Model 3 con olled Model 2 by gende , age, pa en s’ educa ion, he numbe o smoke s among amily membe s, and andom
in e cep a he ci y le el.
*** = P- alue unde 0.001.
‘no- acking’= ampe e ( inland).
Co a ia es Model 1 OR (95%CI) Model 2 OR (95%CI) Model 3 OR (95%CI)
academic pe o mance ( e = high)   
a e age 2.02 (1.74–2.35)*** 1.65 (1.36–2.01)*** 1.55 (1.27–1.90)***
low 3.86 (3.28–4.55)*** 2.26 (1.81–2.84)*** 1.99 (1.58–2.52)***
iendship ies wi h daily smoke s (%)  16.53 (13.36–20.45)*** 13.69 (11.00–17.03)***
iendship ies wi h low pe o me s (%)  1.89 (1.50–2.39)*** 1.76 (1.39–2.23)***
school ype ( e = non- oca ional)   
‘no- acking’  0.96 (0.69–1.34) 1.13 (0.80–1.60)
Voca ional  2.28 (1.87–2.76)*** 1.79 (1.46–2.20)***
age   1.43 (1.32–1.56)***
sex ( e = emale)   
Male   0.98 (0.83–1.15)
Pa en s’ educa ion ( e = high and middle)   
low le el   1.02 (0.85–1.22)
smoking amily membe s   1.48 (1.40–1.56)***
130 P.-O. ROBERT ET AL.
associa ed wi h daily smoking in oca ional schools han in non- oca ional schools and pa icula ly
s ongly in non- acking schools.
Consis ency and in e p e a ion
Adolescen s ended o ha e social ies wi h hose who had he same smoking s a us and academic
pe o mance le el. This is consis en wi h p e ious s udies on bo h smoking (Me cken, Snijde s, S eglich,
& de V ies, 2009; Schae e e al., 2012; Seo & Huang, 2012) and academic pe o mance (Huang e al.,
2014; Kobus, 2003; Lomi e al., 2011; Me cken, Snijde s, S eglich, & de V ies, 2009; Schae e e al., 2012;
Wool e al., 2012). Ou s udy shows ha iendship ne wo ks a e pa e ned on bo h cha ac e is ics
simul aneously. This means ha poo ly pe o ming adolescen s who smoke ha e a double social jeop-
a dy: hey ha e mo e iends who bo h pe o m poo ly and also smoke. Social homophily on hese wo
indi idual cha ac e is ics p o ide he backbone o heal h beha iou inequali y in a school con ex .
One possible explana ion may be ha smoking and low academic pe o mance a e signals o low social
s a us, which may igge g ea e social inclusi eness. Ho n’s s udy (2006), ci ed by Killen, showed ha
adolescen s who iden i ied wi h low-s a us pee -g oup membe s we e socially mo e cohesi e han hei
high-s a us-g oup pee s. This social cohesion, on one hand, ein o ces he social iden i y o a g oup,
bu , on he o he hand, also has consequences o heal h, well-being a school, emo ional engagemen ,
and academic pe o mance (Killen, Ru land, & Jampol, 2009; Ladd, 2009). Ou inding ha ecip ocal
ies we e mo e associa ed wi h he same smoking beha iou and he same le el o pe o mance han
non- ecip ocal ies makes clea he ole o social cohesion (Fujimo o & Valen e, 2012).
This may be ela ed o he s onge associa ion o academic pe o mance and smoking in non- oca-
ional schools (o in schools wi h no acking) han in oca ional schools. Wi hin non- oca ional schools,
s uden s who pe o m wo se a e less likely o ecei e social, emo ional, and academic suppo om
o he s and a e hus mo e likely o clus e oge he and o engage in de ian beha iou such as iola ion
o school ules and smoking (Chen & Hsiao, 2009). The associa ion be ween smoking and academic
pe o mance may be weake in oca ional schools han in non- oca ional schools because, he e, low
academic pe o mance and smoking a e mo e p e alen and hus socially less salien .
Po en ial limi a ions
Ou s udy has some limi a ions. Fi s , he di e ences be ween coun ies’ educa ional sys ems and g ad-
ing sys ems a e likely o a ec he o e all associa ion be ween smoking and academic pe o mance.
In addi ion, he alidi y o he associa ion be ween academic pe o mance and smoking elies on he
alidi y o academic pe o mance, which was sel - epo ed: i is possible ha he social desi abili y o
highe academic pe o mance may lowe he ue associa ion.
Second, ou c oss-sec ional design canno asce ain he di ec ion o he associa ion be ween smok-
ing and academic pe o mance and canno disen angle causal e ec s om selec ion e ec s in he
iendship ne wo k: adolescen s may ha e simila iends in e ms o smoking (o academic pe o mance)
ei he because hey we e in luenced by hei iends’ smoking s a us o because hey selec ed hem
acco ding o hei own smoking s a us (Me cken, Snijde s, S eglich, & de V ies, 2009).
Conclusions
Smoking a es a e highe among adolescen s wi h poo academic pe o mance han among hose wi h
high academic pe o mance. This associa ion was o a conside able ex en explained by homophily
o iendship ies o smoke s and o poo pe o me s. This sugges s ha he educa ional en i onmen
con ibu es o socio-economic inequali ies in smoking in young people. Fu he s udies a e needed o
asce ain whe he educing he clus e ing o poo pe o me s, e.g. class o ma ion, could educe his
associa ion (Valen e, 2012).
INTERNATIONAL JOURNAL OF ADOLESCENCE AND YOUTH 131
Au ho s’ con ibu ion
POR concei ed he esea ch ques ion, pa icipa ed in he da a collec ion, pe o med he da a analysis,
and d a ed he manusc ip . AK con ibu ed o he su ey design and coo dina ed he s udy. POR, MK,
KR, IM, JK, AR, JP, BF, and MR ca ied ou he s udy in each Eu opean ci y and con ibu ed o he man-
usc ip . AK, MK, KR, IM, JK, AR, and JP helped d a he manusc ip . VL designed he su ey, supe ised
he da a collec ion, assis ed in he s udy’s concep ion, and con ibu ed o he manusc ip . All au ho s
ead and app o ed he inal manusc ip .
E hics app o al
In each coun y, e hical app o al om local o na ional e hical commi ees was eques ed and ob ained.
In addi ion, in some coun ies, pe mission o conduc he su ey was eques ed om educa ional
au ho i ies. School p incipals, pa en s, and adolescen s ecei ed lea le s, in o ma ion le e s, and pa en-
al consen le e s, acco ding o each coun y’s egula ions. In o med consen was ob ained om all
indi idual pa icipan s included in he s udy.
I aly
Name o he commi ee: E hics commi ee, 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
Ne he lands
Name o he commi ee: Medical E hical Commi ee o he AMC App o al’s e e ence numbe :
W12_256#12.17.0290
Finland
Name o he commi ee: E hics Commi ee o he Tampe e egion Fa ou able S a emen e e ence
numbe : 10/2012
Ge many
Name o he commi ee: E hics commi ee, Medical Facul y, Ma in-Lu he -Uni e si y Halle-Wi enbe g,
Ge manyApp o al’s e e ence numbe : 2012–112, app o ed on 13/12/2012.
Belgium
Name o he commi ee: Commission d’É hique Biomédicale
App o al’s e e ence numbe : 2012/09OCT/461
Po ugal
Name o he commi ee: Gene al Di ec o a e o Educa ion
(Di eção Ge al da Educação)
App o al’s e e ence numbe : Re numbe 0338600001, app o ed on 02/11/2012
132 P.-O. ROBERT ET AL.