.........................................................................................................
The Eu opean Jou nal o Public Heal h, Vol. 29, No. 1, 44–49
ßThe Au ho (s) 2018. Published by Ox o d Uni e si y P ess on behal o he Eu opean Public Heal h Associa ion.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion Non-Comme cial License
(h p://c ea i ecommons.o g/licenses/by-nc/4.0/), which pe mi s non-comme cial e-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. Fo comme cial e-use, please con ac [email p o ec ed]
doi:10.1093/eu pub/cky179 Ad ance Access published on 4 Sep embe 2018
.........................................................................................................
Low schoolwo k engagemen and schoolwo k
di icul ies p edic smoking in adolescence?
Jaana L. Minkkinen
1
, Jaana M. Kinnunen
1
, Saka i Ka onen
2
, Ris o H. Ho ulainen
3
,
Pi jo L. Lind o s
1,4
, A ja H. Rimpela
¨
1,4,5
1 Facul y o Social Sciences, Uni e si y o Tampe e, Tampe e, Finland
2 Depa men o Heal h and Social Ca e Sys ems, Social Policy Resea ch, Na ional Ins i u e o Heal h and Wel a e,
Helsinki, Finland
3 Depa men o Teache Educa ion, Uni e si y o Helsinki, Helsinki, Finland
4 Tampe e Cen e o Childhood (PERLA), You h and Family Resea ch, Uni e si y o Tampe e, Tampe e, Finland
5 Depa men o Adolescen Psychia y, Pi ka
¨niemi Hospi al, Tampe e Uni e si y Hospi al, Nokia, Finland
Co espondence: Jaana L. Minkkinen, Facul y o Social Sciences (Psychology), Uni e si y o Tampe e, FI-33014 Uni e si y o
Tampe e, Finland, Tel: +358 50 318 7671, Fax: +358 3 213 4473, e-mail: [email p o ec ed]
Backg ound: Low academic achie emen has been associa ed wi h smoking bu ac o s behind his associa ion a e
poo ly known. Such ac o s could include schoolwo k disengagemen and schoolwo k di icul ies. To assess he
ex en o which hey con ibu e o he explana ion o how heal h inequali ies eme ge, we s udy in a longi udinal
design whe he hese ha e an independen e ec on smoking o whe he hei e ec is media ed h ough
academic achie emen . Me hods: Longi udinal da a we e collec ed in he Helsinki me opoli an a ea, Finland in
2011 and 2014. Pa icipan s we e se en h-g ade s (12 13 yea s, N=9497). In he ollow-up, 6534 s uden s epo ed
hei smoking s a us in he nin h g ade (15 16 yea s). Smoking, schoolwo k beha iou al engagemen , i.e. pa ici-
pa ion in academic ac i i ies, and disengagemen , schoolwo k di icul ies and cogni i e compe ence we e sel -
epo ed by adolescen s. Academic achie emen was ob ained om he Finnish na ional applica ion egis e on
uppe seconda y educa ion. A media ion analysis was execu ed wi h boo s apped con idence in e als. Resul s:
Highe schoolwo k beha iou al engagemen and cogni i e compe ence in he se en h g ade p edic ed ha ado-
lescen s we e mo e likely no o smoke in he nin h g ade (all P<0.001) while highe schoolwo k disengagemen
and schoolwo k di icul ies p edic ed adolescen s’ smoking (all P<0.001). The e ec s we e media ed h ough
academic achie emen . Conclusions: S uden s’ beha iou al disengagemen wi h schoolwo k and schoolwo k
di icul ies a e isks o smoking ini ia ion. Thei e ec is media ed h ough poo school achie emen . As
smoking o en con inues in adul hood and poo school pe o mance ypically leads o lowe educa ion,
schoolwo k disengagemen and di icul ies in adolescence cons i u e po en ial pa hways o inequali ies in heal h.
.........................................................................................................
In oduc ion
Socio-economic heal h di e ences in adul hood a e well-
es ablished. People wi h lowe educa ion and socio-economic
posi ion ha e highe mo ali y and mo bidi y compa ed o hose
wi h highe educa ion and socio-economic posi ion.
1
The o igin o
hese heal h di e ences can be de ec ed, a leas pa ly, in adoles-
cence whe e low academic achie emen is ela ed o heal h-
comp omising beha iou s, like smoking. S uden s who achie e
be e g ades a school a e less likely o smoke, while hose whose
pe o mance a school is poo end o smoke mo e o en.
2–6
On he
o he hand, smoking adolescen s comple e ewe yea s o educa ion
compa ed o non-smoke s,
2,7
and s uden s wi h poo academic pe -
o mance ha e a highe p obabili y o ending up wi h a lowe le el o
educa ion in adul hood.
8,9
Added up, his implies ha adolescen s
wi h heal h-comp omising beha iou s a e mo e likely o end up
wi h a low socio-economic posi ion in adul hood han hose wi h
heal hy beha iou s. E en hough he associa ion be ween educa-
ional achie emen and smoking in young people has been shown
in se e al s udies, ac o s behind i a e poo ly known, wi h he
excep ion o low socio-economic backg ound o he amily.
S uden s om lowe socioeconomic backg ound end o smoke
mo e han hose om highe socio-economic backg ound,
10
and a
low socio-economic backg ound p edic s lowe academic achie e-
men in school.
11
Educa ion and heal h ha e mainly been sepa a e ields in he
scien i ic li e a u e, which is why ac o s behind academic achie emen
and isk ac o s o smoking ha e been s udied sepa a ely. In his pape ,
we combine hese wo ields by assessing whe he smoking can be be e
unde s ood wi h concep s de i ed om educa ion esea ch, namely,
schoolwo k engagemen and cogni i e compe ence om lea ning- o-
lea n s udies.
12
We sugges ha hese ac o s may explain he associa ion
be ween smoking and academic achie emen and, consequen ly,
con ibu e o he explana ion o how heal h inequali ies eme ge.
The concep o schoolwo k engagemen comp ises beha iou al,
emo ional and cogni i e dimensions.
13–15
Beha iou al engagemen
has been he dimension ha bes p edic s academic achie emen
16
why his is ou in e es he e. I e e s o in ol emen in lea ning and
academic asks, adhe ing o class oom no ms and pa icipa ion in
academic and ex acu icula ac i i ies.
15
P e ious smoking esea ch
has s udied schoolwo k engagemen in gene al bu he speci ic
con ibu ion o beha iou al engagemen o schoolwo k has been
neglec ed. In hose s udies, schoolwo k engagemen has been
posi i ely associa ed wi h academic achie emen
6,13,17
bu
nega i ely wi h heal h-comp omising beha iou s, like smoking.
6
S uden s’ engagemen o school builds a mo i a ional con ex o
schoolwo k,
18
and enables s uden s o deal wi h di icul ies and ailu es
in school.
19
Schoolwo k di icul ies e e o di icul ies in basic asks,
e.g. asks demanding eading, w i ing o independen wo k. Di icul ies
in hese undamen al skills may wo sen academic achie emen , and
44 Eu opean Jou nal o Public Heal h
Downloaded om h ps://academic.oup.com/eu pub/a icle-abs ac /29/1/44/5090537 by Tampe e Uni e si y and Uni e si y Hospi al use on 29 Janua y 2019
h ough de e io a ion o academic achie emen , schoolwo k
di icul ies may con ibu e o smoking ini ia ion. Addi ionally,
cogni i e skillsp edic academicachie emen ,bu a ecen longi udinal
s udy o ea ly adolescence also showed ha s uden s wi h poo
cogni i e skills end o engage mo e in heal h-comp omising
beha iou s, like smoking, han s uden s wi h good cogni i e skills.
20
We aim o s udy whe he schoolwo k beha iou al engagemen
(in ol emen in lea ning and academic asks), schoolwo k di icul ies
and cogni i e compe ence in he se en h g ade (age 1213 yea s)
p edic ini ia ion o smoking and academic achie emen by he end
o he nin h g ade (age 1516 yea s). Fu he , we s udy whe he he
associa ions be ween smoking and schoolwo k beha iou al
engagemen , schoolwo k di icul ies and cogni i e compe ence a e
media ed h ough academic achie emen . Figu e 1 p esen s ou
s udy design and model.
Me hods
Da a
The longi udinal da a we e collec ed in 2011 and 2014 om all s uden s
in he Helsinki Me opoli an a ea. Pa icipan s in he baseline da a
(se en h-g ade s, 1213 yea s) had jus s a ed lowe seconda y
school (N= 9497; esponse a e 73%; 50.4% gi ls). O he o iginal
coho 6880 s uden s (1516yea s) pa icipa ed in he ollow-up in
he nin h g ade a he end o compulso y school ( esponse a e 72%;
53% o he o iginal sample). Educa ional au ho i ies o 14 munici-
pali ies ga e pe mission o he s udy. The p o ocol was app o ed by
he E hics Commi ee o he Na ional Ins i u e o Heal h and Wel a e.
The baseline and ollow-up su eys included wo ques ionnai es
(Heal h Su ey and Lea ning- o-Lea n Assessmen ). The inal sample
included only s uden s who epo ed hei smoking s a us in he nin h
g ade (N= 6534; 50.4% gi ls, able 1). S uden s’ g ades in he school-
lea ing ce i ica e we e ob ained om he na ional join applica ion
egis e on uppe seconda y educa ion, 2014.
We compa ed he inal sample o hose s uden s who comple ed he
ques ionnai e in he se en h g ade using - es . The longi udinal
sample included less s uden s who had ied smoking in se en h
g ade (P< 0.01), whose pa en s ha e been smoke s (P<0.05),and
who had schoolwo k di icul ies (P< 0.01), and mo e s uden s wi h
highe nin h g ade academic achie emen (P< 0.001) and beha iou al
engagemen in schoolwo k (P<0.05), han hewholesampleo
s uden s, who comple ed he ques ionnai e in 2011. No signi ican
di e ence was ound in o he analysis a iables.
Measu es
Smoking s a us was asked: ‘Which op ion bes desc ibes you
smoking nowadays?’ Response scale was ‘I do no smoke a all’,
‘Less han once a week’, ‘Once a week o mo e o en, bu no
daily’, ‘Once a day o mo e o en’. No smoking was encoded as 0,
o he op ions as 1. In he nin h g ade, 20% smoked a leas
some imes (N= 1309).
Table 1 Sample cha ac e is ics (N
boys
=3255, N
gi ls
=3279)
Boys Gi ls Boys Gi ls Scale T- es
% % Mean SD Mean SD
Dicho omous a iables
Smoking (nin h g ade) 0/1 3.27

Non-smoke s 78.3 81.6
Occasional o daily smoke s 21.7 18.4
Smoking expe ience (se en h g ade) 0/1 6.68

No 78.1 85.2
Yes 21.9 14.8
Pa en s’ smoking 0/1 0.34
No 58.2 58.7
Yes 41.8 41.3
Pa en s’ educa ion 0/1 0.52
O he 63.7 63.1
Uni e si y deg ee 36.3 36.9
Con inuous a iables
a
Academic achie emen (nin h g ade) 7.82 1.06 8.29 1.03 4–10 7.47

Schoolwo k beha iou al engagemen (se en h g ade) 30.78 6.59 32.36 6.27 6–42 10.06

Schoolwo k beha iou al disengagemen (se en h g ade) 13.45 4.08 11.62 4.00 3–21 0.393

Schoolwo k di icul ies (se en h g ade) 9.61 3.18 9.13 2.54 8–24 108.25

Cogni i e compe ence (se en h g ade) 38.55 25.34 41.19 25.22 0–100 0.164

a: Highe sco es e lec highe quan i y o he a ibu e.
:P<0.01, wo- ailed.
:P<0.001, wo- ailed.
Figu e 1 The concep ual model o smoking s a us, academic
achie emen , school- ela ed p edic o s and co a ia es. The s onge
a ows om he a iables in se en h g ade e e o he
hypo hesized o al e ec s h ough he media o o academic
achie emen in nin h g ade on smoking s a us in nin h g ade. A
plus sign e e s o he hypo hesized posi i e associa ion and a
minus sign e e s o hypo hesized nega i e associa ion be ween
a iables
Low schoolwo k engagemen and schoolwo k di icul ies p edic smoking in adolescence? 45
Downloaded om h ps://academic.oup.com/eu pub/a icle-abs ac /29/1/44/5090537 by Tampe e Uni e si y and Uni e si y Hospi al use on 29 Janua y 2019
Academic achie emen was a sum o mo he ongue, ma hema ics,
and o eign language in he school-lea ing ce i ica e. The scale was
4–10 (4= ail, 10= excellen ).
Schoolwo k beha iou al engagemen is ypically desc ibed as
including s uden ’s posi i e conduc , e o on schoolwo k, pay
a en ion in class and p epa e o class.
14,21–22
I was measu ed by
six i ems ega ding s uden ’s homewo k and beha iou du ing
lessons (‘I always ca e ully do he homewo k we ha e been gi en’,
‘I o en lea e my homewo k undone’, ‘I he eache gi es us
homewo k I always do i in ime’, ‘I always lis en ca e ully when
he eache eaches in class’, ‘In lessons, I always lis en ca e ully when
he eache gi es ins uc ions and new asks’, ‘I always lis en ca e ully
when he eache gi es homewo k’). The scale was 1–7 (1= No ue,
7= Ve y ue). I em sco es we e summed o a o al sco e a e
u ning he scale o he second i em. Highe sco es e lec ed
s uden ’s s onge adhe ence o schoolwo k. The in e nal eliabili y
o he composi e a iable was good (C onbach’s = 0.875).
Schoolwo k beha iou al disengagemen
The s uden s we e asked hei opinion in h ee s a emen s: ‘I y o
ge away wi h doing as li le homewo k as possible’, ‘I ha e no
in e es in doing any hing ex a o school’, ‘I only do he
compulso y wo k o school, no hing mo e’. The scale was 1–7
(1 = No ue, 7 = Ve y ue). The ope abili y and alidi y o he
measu e has been alida ed in educa ional esea ch.
23
I em sco es
we e summed o a o al sco e (C onbach’s = 0.732). Highe sco es
e lec ed s uden ’s s onge schoolwo k beha iou al disengagemen .
Schoolwo k di icul ies, he se en h g ade, we e measu ed using an
eigh -i em indica o which has been used in he Finnish School
Heal h P omo ion s udies.
24
The ques ion was ‘How a e you
doing a school? Do you ha e di icul ies in he ollowing a eas’:
paying a en ion o he eaching du ing class hou s, wo king in
g oups, doing homewo k o simila asks, p epa ing o exams,
inding he igh s udy me hod o you, doing assignmen s ha
equi e you o wo k independen ly, doing assignmen s ha equi e
w i ing, doing assignmen s ha equi e eading (e.g. a book). The
op ions we e on a ou -poin scale: No, Some, Qui e a lo , Ve y
much. The op ions ‘No’ and ‘Some’ we e coded as 1, ‘Qui e a lo ’
as 2 and ‘Ve y much’ as 3. The i ems we e combined o o m a
composi e a iable wi h a highe alue indica ing mo e
schoolwo k di icul ies (C onbach’s = 0.927).
Cogni i e compe ence was cons uc ed by he mean o wo
cogni i e asks, aken om he Finnish Lea ning- o-Lea n-
Assessmen ba e y.
24
Bo h in oduced cogni i e scales ha e been
ecen ly used in di e en esea ch p ojec s using SEM, e idencing
hei ope abili y and alidi y.
23,25,26
The Con ol o Va iables ask
consis ed o eigh i ems, each including a compa ison pai o
a iables in he wo ld o Fo mula 1 aces: D i e , Ca , Ti es and
T ack. One i em included h ee o ou subsec ions whe e he ado-
lescen s had o decide whe he he compa ison pai ga e in o ma-
ion o decide he single e ec o he a iables. The op ions in each
subsec ion we e: ‘No’, ‘Maybe’, and ‘Yes’. All subsec ions had o be
answe ed co ec ly o ge poin s om he i em. The ask sco es we e
pe cen o co ec ly answe ed i ems (1 i em =12.50, 2 i ems =25.00
e c.). The maximum sco e was 100. The eliabili y was adequa e
(C onbach’s = 0.790). The o he cogni i e ask was he In en ed
Ma hema ical Concep s ask, which was a modi ied e sion o
S e nbe g and colleagues’
27
T ia chic Tes C ea i e Numbe
scale.
23
In he ask, wo imagina y ma hema ical concep s we e
in oduced o adolescen s, who we e hen asked o sol e se en
a i hme ic asks (mul iple choice al e na i es) based on he new
concep s. The sco es we e he pe cen age o co ec answe s (1
co ec answe =14.29, 2 co ec answe s = 28.57 e c.). The
maximum sco e was 100. The eliabili y was adequa e (C onbach’s
= 0.750).
Co a ia es
S uden ’s smoking expe ience, pa en s’ smoking and pa en s’
educa ion we e backg ound a iables. Smoking expe ience in he
se en h g ade was measu ed; ‘Ha e you e e smoked? I you ha e,
how many ciga e es ha e you smoked hus a ?’ wi h op ions ‘I ha e
ne e ied o smoke’, ‘1’, ‘Abou 2 50’, ‘Mo e han 50’. ‘Ne e
ied’ was coded as 0, he o he s we e coded as 1. Mo he s’ and
a he s’ smoking was asked sepa a ely: ‘Ha e you pa en s smoked
du ing you li e ime?’ wi h op ions ‘He/she has ne e smoked’,
‘He/she has smoked bu qui ’, ‘He/she smokes nowadays’, ‘I ha e
no a he /mo he ’. Pa en s’ smoking was coded as 1, i ei he pa en
smokes nowadays o had smoked bu qui . ‘Ne e smoked’ and ‘No
mo he / a he ’ we e coded as 0. Pa en s’ educa ion was asked
sepa a ely o mo he s and a he s: ‘Wha kind o educa ion do
you pa en s ha e?’ wi h op ion ‘Basic educa ion only’, ‘Voca ional
uppe seconda y educa ion o oca ional college’, ‘Ma icula ion
examina ion and oca ional college’ ‘Uni e si y deg ee’, ‘No
mo he / a he ’. The highes le el o ei he pa en ’s educa ion was
used. Pa en al educa ion was dicho omized o he echnical
easons as co a ia es should be dicho omic o con inuous in he
media ion analysis using he Mplus s a is ical package.
28
Uni e si y
deg ee was encoded as 1, o he op ions as 0. ‘No mo he and a he ’
was coded as missing. O s uden s, 37% had a leas one pa en wi h
uni e si y-le el educa ion which equa es he high educa ion le el in
he popula ion o he Helsinki Me opoli an a ea.
29
S a is ical analysis
Desc ip i e and p elimina y analyses we e pe o med wi h IBM SPSS
s a is ics 23 and he pa h analysis was execu ed using he Mplus
( e sion 8) applying he WLSMV es ima o .
28
The pa h analysis was
accomplished in o de o explo e he media ed e ec s o he a iables
measu ed in he se en h g ade on smoking s a us ( he nin h g ade)
h ough academic achie emen ( he nin h g ade; igu e 1). The
media ion model analyses we e conduc ed sepa a ely o boys and
gi ls, as he p elimina y analyses showed ha boys and gi ls di e ed
signi ican ly in he mos o dependen a iables ( able 1). All a iables
measu ed in he se en h g ade we e allowed o co ela e wi h each
o he . Boo s apped con idence in e als we e conduc ed o
de e mine eliable con idence in e als o indi ec e ec s, which a e
no no mally dis ibu ed in na u e.
30
To al, indi ec and di ec e ec s
we e based on coun e ac uals, which akes in o accoun ha he
ou come is bina y and he media o is con inuous.
30
Odds a ios
o smoking in he nin h g ade we e epo ed. Two- ailed signi icance
es ing was applied a he c i e ion le el o P= 0.05. Missing da a
anged om 1.4 o 28.7%. Missing da a we e handled in Mplus using
a ull in o ma ion maximum likelihood es ima ion.
28
Resul s
The causally-de ined di ec and o al e ec s o he analyzed a i-
ables we e all signi ican on he smoking s a us among bo h
gende s (all P< 0.001; able 2). The di ec and o al e ec s
showed ha highe schoolwo k beha iou al engagemen in he
se en h g ade p edic ed less smoking in bo h gende s in he nin h
g ade (di ec e ec boys OR = 0.717, gi ls OR = 0.710; o al e ec
boys OR = 0.670; gi ls OR = 0.655). Con e sely, highe schoolwo k
beha iou al disengagemen p edic ed mo e smoking in he nin h
g ade (di ec e ec boys OR= 1.099, gi ls OR = 1.072; o al e ec
boys OR = 1.113, gi ls OR = 1.108). Addi ionally, schoolwo k
di icul ies p edic ed mo e smoking (di ec e ec boys OR =
1.004, gi ls OR = 1.102; o al e ec boys OR = 1.030, OR =
1.141). Highe cogni i e compe ence in he se en h g ade
p edic ed less smoking in he nin h g ade among boys and gi ls
acco ding o he o al e ec ( espec i ely, OR = 0.784, OR =
0.903). In sum, he hypo hesized longi udinal associa ions we e
con i med acco ding o he o al e ec s. Tha is, he e ec s o
46 Eu opean Jou nal o Public Heal h
Downloaded om h ps://academic.oup.com/eu pub/a icle-abs ac /29/1/44/5090537 by Tampe e Uni e si y and Uni e si y Hospi al use on 29 Janua y 2019
s uden s’ schoolwo k beha iou al engagemen , schoolwo k beha -
iou al disengagemen , schoolwo k di icul ies and cogni i e
compe ence on smoking we e media ed h ough academic
achie emen .
All con olling a iables, adolescen ’s smoking expe ience,
pa en s’ smoking and pa en s’ educa ion had signi ican di ec and
o al e ec s on boys’ and gi ls’ smoking (all P< 0.001; able 2).
All he a iables measu ed in he se en h g ade signi ican ly
co ela ed wi h each o he in boys (P< 0.01) and gi ls (P<
0.001). R
2
o smoking s a us was 0.252 in boys and 0.289 in gi ls,
and o academic achie emen 0.401 o bo h gende s.
Discussion
We in es iga ed he longi udinal associa ions be ween s u-
den s’ ini ia ion o smoking and hei beha iou al engagemen
in schoolwo k and schoolwo k di icul ies. Ou esul s go
beyond he known ela ionship be ween poo academic
Table 2 P edic ing smoking s a us in he nin h g ade h ough academic achie emen (N=6534)
Boys Gi ls
OR SE OR SE
Va iables in se en h g ade
Schoolwo k beha iou al engagemen
To al na u al indi ec e ec
a
0.934

0.012 0.922

0.014
Pu e na u al di ec e ec
a
0.717

0.044 0.710

0.046
To al e ec
a
0.670

0.042 0.655

0.044
Schoolwo k beha iou al disengagemen
To al na u al indi ec e ec 1.013

0.008 1.034

0.013
Pu e na u al di ec e ec 1.099

0.069 1.072

0.068
To al e ec 1.113

0.070 1.108

0.071
Schoolwo k di icul ies
To al na u al indi ec e ec 1.026

0.008 1.036

0.012
Pu e na u al di ec e ec 1.004

0.052 1.102

0.056
To al e ec 1.030

0.052 1.141

0.059
Cogni i e compe ence
To al na u al indi ec e ec 0.822

0.024 0.759

0.022
Pu e na u al di ec e ec 0.954

0.062 1.189

0.073
To al e ec 0.784

0.045 0.903

0.051
Co a ia es
Smoking expe ience
To al na u al indi ec e ec 1.057

0.011 1.079

0.015
Pu e na u al di ec e ec 1.519

0.069 1.701

0.083
To al e ec 1.606

0.072 1.835

0.089
Pa en s’ smoking
To al na u al indi ec e ec 1.027

0.008 1.023

0.010
Pu e na u al di ec e ec 1.240

0.065 1.246

0.067
To al e ec 1.273

0.067 1.274

0.069
Pa en s’ educa ion
To al na u al indi ec e ec 0.945

0.010 0.913

0.011
Pu e na u al di ec e ec 0.971

0.048 1.101

0.057
To al e ec 0.918

0.045 1.005

0.052
Boys Gi ls
bSEBSE
E ec o academic achie emen in nin h
g ade o smoking in nin h g ade
0.245

0.037 0.364

0.040
E ec s o a iables in se en h g ade o academic achie emen in nin h g ade
Schoolwo k beha iou al engagemen 0.150

0.017 0.122

0.020
Schoolwo k beha iou al disengagemen 0.030 0.017 0.054

0.018
Schoolwo k di icul ies 0.059

0.016 0.056

0.018
Cogni i e compe ence 0.444

0.015 0.439

0.015
Co a ia es
Smoking expe ience 0.136

0.016 0.129

0.019
Pa en s’ smoking 0.063

0.017 0.036

0.015
Pa en s’ educa ion 0.130

0.016 0.146

0.014
R
2
Smoking s a us in nin h g ade 0.401 0.401
Academic achie emen in nin h g ade 0.252 0.289
Model i
Numbe o ee pa ame e s 53 53
RMSEA 0.000 0.000
CFI/TLI 1.000/1.000 1.000/1.000
N3255 3279
a: To al, indi ec , and di ec e ec s based on coun e ac uals. OR = odds a io. Highe sco es e lec highe quan i y o he a ibu e.
No e: b=uns anda dized coe icien ; SE = s anda d e o . Highe sco es e lec highe quan i y o he a ibu e. The co ela ions be ween
he a iables in se en h g ade signi ican ly co ela ed wi h each o he bu hey a e no displayed in he able o conse e space.
:P< 0.05, wo- ailed.
:P< 0.01, wo- ailed.
:P< 0.001, wo- ailed.
Low schoolwo k engagemen and schoolwo k di icul ies p edic smoking in adolescence? 47
Downloaded om h ps://academic.oup.com/eu pub/a icle-abs ac /29/1/44/5090537 by Tampe e Uni e si y and Uni e si y Hospi al use on 29 Janua y 2019
achie emen and smoking by showing ha low engagemen in
schoolwo k and schoolwo k di icul ies p edic smoking
and ha hei e ec s media e pa ly h ough academic achie e-
men . Gi en ha adolescen smoking o en pe sis s o adul -
hood and ha poo academic achie emen p edic s sho
educa ion, we ind ha low schoolwo k engagemen and
schoolwo k di icul ies cons i u e a po en ial pa hway o heal h
inequali ies.
Ea ly nega i e a i udes o schoolwo k ep esen ma ke s o an
inc eased isk o smoking ini ia ion du ing he la e phases o ado-
lescence. P e iously, he associa ion be ween lowe schoolwo k
engagemen and adolescen smoking has been ound in a c oss-
sec ional se ing.
6
Ou s udy shows ha he associa ion can also be
es ablished in a longi udinal design, so ha bo h low schoolwo k
beha iou al engagemen and schoolwo k di icul ies p edic la e
adolescen smoking. P e ious s udies ha e no been able o de ec
ha his associa ion is media ed h ough academic achie emen .
Ou s udy also con i med p e ious indings ha s uden s wi h
lowe cogni i e compe ence end o engage mo e in heal h-
comp omising beha iou s han s uden s wi h highe cogni i e
compe ence.
20
The longi udinal associa ion o high schoolwo k be-
ha iou al engagemen and academic achie emen was also de ec ed
in ou s udy consis en ly wi h p e ious s udies, which ha e shown
ha especially highe engagemen in schoolwo k p edic s subsequen
academic achie emen .
16
Ou s udy p o ides new insigh in o he well- epo ed connec ion
be ween poo academic achie emen and smoking. The ollowing
heo e ical in e p e a ions o e explana ions o he associa ion.
The i s is inhe en ly a psychological heo y ha es s on he
no ion o conse a ion o esou ces (COR).
31
The COR heo y
sugges s ha s ess is he eac ion o he h ea o, o loss o ,
aluable psychological, ma e ial and social esou ces, and ha o
elie e o ejec nega i e s ess esponses, people a e in ui i ely
mo i a ed o p o ec , conse e and eco e hese aluable
esou ces.
31
Fo schoolchild en, poo academic achie emen is
de ini ely a s esso . Ac ually, i is an inc easing numbe o
s esso s h ough cyclic assessmen s a school in which a poo ly
pe o ming s uden equen ly inds him/he sel o be wo se
compa ed o his/he pee s. Thus, poo academic achie emen may
gene a e plen y o eelings o loss in s uden s, e.g. loss o pleasu e by
compe ency o loss o eache s’ and pa en s’ app ecia ion. I also
may lead o ma e ial loss, as poo academic achie emen p edic s
low-pay employmen in adul hood. Acco ding o COR, poo ly
pe o ming s uden s end o p o ec , conse e, and eco e hei
loss o psychological and social esou ces.
31
As nico ine induces
pleasu e and educes s ess and anxie y, imp o es concen a ion
and pe o mance o ce ain asks,
32
o poo ly pe o ming s uden s
smoking may become a way o con ol mood and a way o achie e
eco e y om he psychological and social s ess caused by poo
academic achie emen .
Ano he possible explana ion o he associa ion be ween low
schoolwo k beha iou al disengagemen and smoking could de i e
om he sel -sys em model, based on sel -de e mina ion heo y.
18
This sugges s ha he ex en o which s uden s a e engaged o
disengaged in school cons i u es a mo i a ional con ex , which
also enables s uden s o deal wi h di icul ies and ailu es in
school.
19
Consequen ly, low schoolwo k beha iou al engagemen
migh lead o sea ching o com o in heal h-comp omising
beha iou s like smoking. This migh in u n u he inc ease disen-
gagemen o school, as he ela ionship be ween school engagemen
and heal h-comp omising beha iou has been shown o be bi-
di ec ional.
33
One limi a ion o ou s udy is he use o sel - epo ed da a.
The e o e, smoking migh be unde epo ed o o e - epo ed, as
s uden s may ha e answe ed in socially accep able ways in
class ooms.
34
Adolescen s’ sel - epo o smoking has, howe e ,
been ound o be in ag eemen wi h biochemical measu es.
35–37
Ano he limi a ion is ha only s uden s li ing in he me opoli an
a ea o Helsinki cons i u ed he coho , and hey may no ep esen
all Finnish adolescen s. This migh comp omise he gene alizabili y
o he esul s o some ex en , as people in he me opoli an a ea a e
be e educa ed han elsewhe e in Finland.
29,38
The non- esponse
analysis showed some selec ion o he longi udinal sample so ha
s uden s wi h smoking and schoolwo k di icul ies we e
unde ep esen ed and hose wi h be e academic achie emen and
beha iou al engagemen we e o e ep esen ed. This may a ec
s eng hs o he associa ions bu is unlikely o bias hem. The
s eng h o ou s udy is ha we we e able o s udy he associa ions
in a longi udinal se ing allowing us o e eal po en ially causal
pa hways. The coho also consis ed o nume ous s uden s, a
qua e o he en i e age coho in Finland, which allowed o
conduc he analyses wi h espec o gende . Addi ionally, se e al
i ems cons i u ed he a iables wi h good eliabili ies, hus
cap u ing a ious aspec s o he phenomena.
Acknowledgemen s
The esul s ha e p e iously been p esen ed in he 10 h Eu opean
Public Heal h Con e ence, S ockholm, Sweden on 3 No embe 2017.
Funding
This wo k was suppo ed by Academy o Finland, Helsinki, Finland,
in 201013 (g an numbe : 135019) and in 201518 (g an
numbe : 288774), and by he Compe i i e Resea ch Funding o
he Tampe e Uni e si y Hospi al (g an numbe : 9S055), Tampe e,
Finland.
Con lic s o in e es : None decla ed.
Key poin s
S uden s’ low beha iou al engagemen wi h schoolwo k and
schoolwo k di icul ies p edic smoking ini ia ion.
The e ec o low schoolwo k engagemen and schoolwo k
di icul ies on smoking ini ia ion is pa ly media ed h ough
poo school achie emen .
Gi en ha adolescen smoking o en pe sis s o adul hood
and ha poo academic achie emen p edic s sho
educa ion, low schoolwo k engagemen and schoolwo k
di icul ies can be conside ed a po en ial pa hway o heal h
inequali ies.
Suppo o s uden s wi h schoolwo k di icul ies and poo
cogni i e skills could p o ide a bu e agains he ini ia ion
o smoking.
Re e ences
1 Commission on Social De e minan s o Heal h (CSDH). Closing he Gap in a
Gene a ion: Heal h Equi y h ough Ac ion on he Social De e minan s o Heal h.
Final Repo o he Commission on Social De e minan s o Heal h. Gene a,
Swi ze land: Wo ld Heal h O ganiza ion, 2008. A ailable a : h p://www.who.in /
social_de e minan s/ hecommission/ inal epo /en/ (25 Janua y 2018, da e las
accessed).
2 B adley B, G eene A. Do heal h and educa ion agencies in he Uni ed S a es sha e
esponsibili y o academic achie emen and heal h? A e iew o 25 yea s o e idence
abou he ela ionship o adolescen s’ academic achie emen and heal h beha io s.
J Adolesc Heal h 2013;52:523–32.
3 B yan A, Schulenbe g J, Bachman JG, e al. Unde s anding he links among school
misbeha io , academic achie emen , and ciga e e use: a na ional panel s udy o
adolescen s. P e Sci 2000;1:71–87.
48 Eu opean Jou nal o Public Heal h
Downloaded om h ps://academic.oup.com/eu pub/a icle-abs ac /29/1/44/5090537 by Tampe e Uni e si y and Uni e si y Hospi al use on 29 Janua y 2019

4 Pennanen M, Haukkala A, De V ies H, Va iainen E. Longi udinal s udy o ela ions
be ween school achie emen and smoking beha io among seconda y school
s uden s in Finland: esul s o he ESFA s udy. Subs Use Misuse 2011;46:569–79.
5 Schnoh C, K eine S, Rasmussen M, e al. School- ela ed media o s in social
inequali ies in smoking: a compa a i e c oss-sec ional s udy o 20, 399 adolescen s.
In J Equi y Heal h 2009;8:17.
6 Kinnunen J, Lind o s P, Rimpela
¨A. Academic well-being and smoking among 14-
o 17-yea -old schoolchild en in six Eu opean ci ies. J Adolesc 2016;50:56–64.
7 La ala A, Rose R, Pulkkinen L, e al. D inking, smoking, and educa ional
achie emen : c oss-lagged associa ions om adolescence o adul hood. D ug Alcohol
Depend 2014;137:106–13.
8 Koi usil a L, Rimpela
¨A, Vika A. Heal h beha iou s and heal h in adolescence as
p edic o s o educa ional le el in adul hood: a ollow-up s udy om Finland. Soc Sci
Med 2003;57:577–93.
9 Acacio-Cla o P, Doku D, Koi usil a L, Rimpela
¨A. How socioeconomic ci cums ances,
school achie emen and ese e capaci y in adolescence p edic adul educa ion le el: a
h ee-gene a ion s udy in Finland. In J Adolesc You h 2017;23:382–97.
10 Moo I, Ra hmann K, Lenzi M, e al. Socioeconomic inequali ies in adolescen
smoking ac oss 35 coun ies: a mul ile el analysis o he ole o amily, school and
pee s. Eu J Public Heal h 2015;25:457–63.
11 O ganisa ion o Economic Co-ope a ion and De elopmen (OECD). Socio-
economic s a us, s uden pe o mance and s uden s’ a i udes owa ds science. In:
PISA 2015 Resul s (Volume I): Excellence and Equi y in Educa ion. Pa is, F ance:
OECD Publishing, 2016: 201–39.
12 Hau ama
¨ki J, Kupiainen S. Lea ning o lea n in Finland. Theo y and policy, esea ch
and p ac ice. In: Deakin CR, S inghe C, Ren K, edi o s. Lea ning o Lea n.
In e na ional Pe spec i es om Theo y and P ac ice. London: Rou ledge, 2014: 179–205.
13 Upadyaya K, Salmela-A o K. De elopmen o school engagemen in associa ion wi h
academic success and well-being in a ying social con ex s. A e iew o empi ical
esea ch. Eu Psychol 2013;18:136–47.
14 Apple on J, Ch is enson S, Fu long M. S uden engagemen wi h school: c i ical
concep ual and me hodological issues o he cons uc . Psychol Sch 2008;45:369–86.
15 F ed icks J, Blumen eld P, Pa is A. School engagemen : po en ial o he concep ,
s a e o he e idence. Re Educ Res 2004;74:59–109.
16 S e ansson K, Ges sdo i S, Geldho G, e al. A bi ac o model o school
engagemen : assessing gene al and speci ic aspec s o beha io al, emo ional and
cogni i e engagemen among adolescen s. In J Beha De 2016;40:471–80.
17 Salmela-A o K, Upadaya K. The schoolwo k engagemen in en o y: ene gy,
dedica ion, and abso p ion (EDA). Eu J Psychol Assess 2012;28:60–7.
18 Ryan R, Deci E. Wha is he sel in sel -di ec ed lea ning? Findings om ecen mo-
i a ional esea ch. In: S aka G, edi o . Concep ions o Sel -Di ec ed Lea ning: Theo e ical
and Concep ual Conside a ions. Muns e . Ge many: Waxmann, 2000: 75–92.
19 Skinne E, Pi ze J. De elopmen al dynamics o engagemen , coping, and e e yday
esilience. In: Ch is enson S, Reschly A, Wylie C, edi o s. The Handbook o Resea ch
on S uden Engagemen . New Yo k, NY: Sp inge Science, 2012: 21–44.
20 Zilanawala A, Sacke A, Kelly Y. Longi udinal la en cogni i e p o iles and psy-
chosocial well-being in ea ly adolescence. J Adolesc Heal h 2017;61:493–500.
21 Klem AM, Connell JP. Rela ionships ma e : linking eache suppo o s uden
engagemen and achie emen . J School Heal h 2004;74:262–73.
22 Ma ks HM. S uden engagemen in ins uc ional ac i i y: pa e ns in he elemen a y,
middle, and high school yea s. Am Educ Res J 2000;37:153–84.
23 Kupiainen S, Vainikainen M-P, Ma janen J, Hau ama
¨ki J. The ole o ime on ask
in compu e -based low-s akes assessmen o c oss-cu icula skills. J Educ Psychol
2014;106:627–38.
24 Luopa P, Ki ima
¨ki H, Ma ikka A. Nuo en Hy in oin i Suomessa 2000–13.
Koulu e eyskyselyn ulokse [Wellbeing o adolescen s in Finland 2000–2013.
The Resul s o he School Heal h P omo ion s udy]. Na ional Ins i u e o Heal h
and Wel a e (THL). Repo 25/2014. Helsinki, Finland. A ailable a : h p://www.
julka i. i/bi s eam/handle/10024/116692/URN_ISBN_978-952-302-280-5.pd ?
sequence=1&isAllowed=y (26 Juny 2018, da e las accessed).
25 Vainikainen M-P, Hau ama
¨ki J, Ho ulainen R, Kupiainen S. Gene al and speci ic
hinking skills and schooling: p epa ing he mind o new lea ning. Think Skills C ea
2015;18:53–64.
26 Vainikainen M-P, Hienonen N, Hau ama
¨ki J, Ho ulainen R. Tukea a i se ien
oppilaiden aja elu ai ojen kehi yminen e ikokoisissa luokissa [De elopmen o
hinking skills o SEN s uden s and class size.]. NMI Bull 2015;25:36–51.
27 S e nbe g R, Cas ejon J, P ie o M, e al. Con i ma o y ac o analysis o he
S e nbe g T ia chic Abili ies Tes in h ee in e na ional samples. Eu J Psychol Assess
2001;17:1–16.
28 Mu he
´n LK, Mu he
´n BO. Mplus Use ’s Guide, 7 h edn. Los Angeles, CA: Mu he
´n&
Mu he
´n, 1998–2012.
29 O icial S a is ics o Finland (OSF). Va
¨es o
¨n koulu us akenne 2012 [Educa ional
s uc u e o he popula ion in 2012]. Appendix igu e 4. 20 uo a a
¨y a
¨neen
a
¨es o
¨n koulu us aso kunni ain 2012 (ka a) [Educa ional le el o he 20-yea -old
popula ion in municipali ies in 2012 (a map)]. Helsinki: S a is ics Finland. A ailable
a : h p://www.s a . i/ il/ kou /2012/ kou _2012_2013-12-04_ku _004_ i.h ml
(25 Oc obe 2017, da e las accessed).
30 Mu he
´n B, Mu he
´n L, Aspa ouho T. Reg ession and Media ion Analysis Using
Mplus. Los Angeles, CA: Mu he
´n & Mu he
´n, 2016.
31 Hob oll S. Conse a ion o esou ces. A new a emp a concep ualizing s ess. Am
Psychol 1989;44:513–24.
32 Benowi z N. Nico ine addic ion. N Engl J Med 2010;362:2295–303.
33 Wang M-T, F ed icks J. The ecip ocal links be ween school engagemen , you h
p oblem beha io s, and school d opou du ing adolescence. Child De 2014;85:722–37.
34 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.
35 Dolcini M, Adle N, Lee P, Bauman K. An assessmen o he alidi y o adolescen sel -
epo ed smoking using h ee biological indica o s. Nico ine Tob Res 2003;5:473–83.
36 Ken ala J, U iainen P, Pahkala K, Ma ila K. Ve i ica ion o adolescen sel - epo ed
smoking. Addic Beha 2004;29:405–11.
37 Pos A, Gilljam H, Rosendahl I, e al. Validi y o sel epo s in a coho o
Swedish adolescen smoke s and smokeless obacco (snus) use s. Tob Con ol
2005;14:114–7.
38 O icial S a is ics o Finland (OSF). Tulonjaon kokonais ilas o [Income dis ibu ion
s a is ics]. Tulokehi ys aluei ain 2015 [A eal income de elopmen 2015], 2.
Pieni uloisia kun ia keski yny keskisen Pa
¨ija
¨n een seudulle [2. Low-income
municipali ies concen a ed in he Pa
¨ija
¨nne a ea]. Helsinki: S a is ics Finland.
A ailable a : h p://www.s a . i/ il/ jk /2015/01/ jk _2015_01_2016-12-20_ka _002_
i.h ml (25 Oc obe 2017, da e las accessed).
Low schoolwo k engagemen and schoolwo k di icul ies p edic smoking in adolescence? 49
Downloaded om h ps://academic.oup.com/eu pub/a icle-abs ac /29/1/44/5090537 by Tampe e Uni e si y and Uni e si y Hospi al use on 29 Janua y 2019