Con en s lis s a ailable a ScienceDi ec
Jou nal o Psychosoma ic Resea ch
jou nal homepage: www.else ie .com/loca e/jpsycho es
Change in β
2
-agonis use a e se e e li e e en s in adul s wi h as hma: A
popula ion-based coho s udy
Li e e en s and b onchodila o usage among adul s wi h as hma
Raija Lie zén
a,⁎
, Pekka Vi anen
b
, Mika Ki imäki
c,d
, Jy ki Ko keila
e
, Saka i Suominen
a,
,
Lau i Sillanmäki
c
, Ma kku Kosken uo
c
, Jussi Vah e a
g
a
Depa men o Public Heal h, Uni e si y o Tu ku, Tu ku, Finland
b
School o Heal h Sciences, Uni e si y o Tampe e, Tampe e, Finland
c
Depa men o Public Heal h, Clinicum, Facul y o Medicine, Uni e si y o Helsinki, Helsinki, Finland
d
Depa men o Epidemiology and Public Heal h, Uni e si y College London Medical School, London, Uni ed Kingdom
e
Depa men o Psychia y, Uni e si y o Tu ku and Ha ja al a Hospi al, Sa akun a Hospi al Dis ic , Ha ja al a, Finland
Depa men o Public Heal h, Uni e si y o Skö de, Skö de, Sweden
g
Depa men o Public Heal h, Uni e si y o Tu ku and Tu ku Uni e si y Hospi al, Tu ku, Finland
ARTICLE INFO
Keywo ds:
As hma
Coho s udy
Gene alized es ima ing equa ion me hod
Li e e en s
Sho -ac ing β
2
-agonis s
S ess ul li e e en s
ABSTRACT
Objec i e: This p ospec i e, popula ion-based coho s udy o 1102 Finnish adul s wi h as hma, examined
whe he exposu e o s ess ul li e e en s is associa ed wi h he in ensi y o usage o inhaled sho -ac ing β
2
-
agonis s.
Me hods: Su ey da a was collec ed by wo pos al ques ionnai es. Baseline cha ac e is ics we e ob ained in 1998
and da a on 19 specific s ess ul e en s (e.g. dea h o a child o spouse o di o ce) wi hin he six p eceding
mon hs in 2003. Exposu e o li e e en s was indica ed by a sum sco e weigh ed by mean se e i y o he e en s.
Pa icipan s we e linked o eco ds o filled p esc ip ions o inhaled sho -ac ing β
2
-agonis s om na ional
egis e s om 2000 h ough 2006. The a es o pu chases o sho -ac ing β
2
-agonis s be o e (2000−2001),
du ing (2002−2003) and a e (2004–2006) he e en exposu e we e es ima ed using epea ed-measu es
Poisson eg ession analyses wi h he gene alized es ima ing equa ion.
Resul s: O he 1102 pa icipan s, 162 (15%) we e exposed o highly s ess ul e en s, 205 (19%) o less s ess ul
e en s. Du ing he 7-yea obse a ion pe iod, 5955 pu chases o filled p esc ip ion o inhaled sho -ac ing β
2
-
agonis s we e eco ded. A e exposu e o highly s ess ul e en s, he a e o pu chases o β
2
-agonis s was 1.50
imes highe (95% confidence in e al (CI): 1.05, 2.13) han be o e he s ess ul e en occu ed. Among hose
wi h low o no exposu e o li e e en s, he co esponding a e a ios we e no ele a ed ( a e a io 0.81, 95% CI:
0.66, 0.99 and 0.95, 95% CI: 0.83, 1.09 espec i ely).
Conclusion: An inc ease in β
2
-agonis usage a e se e e li e e en s sugges s ha s ess ul expe iences may
wo sen as hma symp oms.
1. In oduc ion
As hma is a ch onic in e mi en inflamma ion o he la ge ai ways
[1,2] wi h he epo ed popula ion p e alence a ying om 2% in Es-
onia o 21% in Aus alia [3]. I s p e alence is inc easing in many
coun ies [4].A la ge numbe o s udies on he biological isk ac o s o
as hma mo bidi y ha e ound e idence o he e iological impo ance o
espi a o y in ec ions, alle gens, ai pollu an s, and obacco smoke
[5,6]. Recen ly, he ole o psychosocial s ess as a con ibu o o
as hma mo bidi y has gained inc eased a en ion [7,8]. S ess is
conside ed o affec he exace ba ion o as hma h ough mul iple im-
mune, endoc ine, neu al, and beha iou al p ocesses [9,10]. S ess also
accen ua es he indi idual's immune esponse and induces changes in
inflamma o y p ocesses in he ai ways [11].
Some longi udinal coho s udies sugges an associa ion be ween
nega i e s ess ul li e e en s in he amily and elsewhe e and as hma
onse [12–14], while o he s ha e no ound any associa ion [15]. The
mos ecen s udy sugges s ha bo h wo k s ess and amily ela ed li e
e en s a e posi i ely associa ed wi h as hma in women [16]. In addi-
ion, li e e en s may lead o a wo sening o as hma among as hma ic
h p://dx.doi.o g/10.1016/j.jpsycho es.2017.07.003
Recei ed 16 Feb ua y 2017; Recei ed in e ised o m 15 June 2017; Accep ed 4 July 2017
⁎
Co esponding au ho a : Depa men o Public Heal h, Uni e si y o Tu ku, FIN-20014 Tu un yliopis o, Finland.
E-mail add ess: aija.lie zen@u u.fi(R. Lie zén).
Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
0022-3999/ © 2017 The Au ho s. Published by Else ie Inc. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/BY/4.0/).
MARK
adul s [17–19]. Li e e en s we e isk ac o s o hospi al admissions due
o as hma in a popula ion based s udy [18] and hospi al admission o
acu e se e e as hma in a case-con ol s udy [19]. Exposu e o com-
muni y iolence was associa ed wi h bo h hospi al admission and
eme gency depa men isi s due o as hma in a s udy among adul s
[17] while long-las ing s ess, a leas a wo k, was no associa ed wi h
se e e as hma exace ba ions leading o hospi aliza ion o dea h [20].
The p e ious s udies scope on hospi aliza ions, which ep esen se ious
exace ba ions o as hma ha occu a imes in mos pa ien s and ha e
dec eased du ing he las yea s [21].
P esc ip ions o as hma medica ion may p o ide a way o assessing
day- o-day a ia ion in e e yday as hma symp oms in ela ion o in-
c ease in li e s ess. To da e, howe e , he associa ion be ween ecen
s ess ul li e e en s and usage o inhaled b onchodila o s has no been
s udied in as hma ic adul s in spi e o he ac ha a co ne s one o
success ul ea men o as hma is sel -managemen . I includes he
pe cep ion o as hma symp oms and he use o p esc ibed medica ion.
The aim o medica ion is o con ol he disease and o p e en i s ex-
ace ba ion. Inhaled co icos e oids, in combina ion wi h o sepa a e
om long-ac ing β
2
-agonis s as an i-inflamma o y d ugs, play a majo
ole in his ea men . In addi ion, inhaled sho -ac ing β
2
-agonis s a e
used o b onchodila ion and p o ec ion agains b onchocons ic ion
and as quick- elie d ugs o as hma symp oms. [22,23]
In his p ospec i e s udy, we hypo hesized ha people wi h as hma
inhale sho -ac ing β
2
-agonis s when hey a e symp oma ic and ha
high exposu e o ecen s ess ul li e e en s would be associa ed wi h
wo sening symp oms, esul ing in inc eased pu chase o inhaled sho -
ac ing β
2−
-agonis medica ion. To examine his, we used da a o filled
p esc ip ions o as hma ic pe sons be o e, du ing, and a e exposu e
o ecen s ess ul li e e en s o e a se en-yea obse a ion pe iod. To
he bes o ou knowledge, no la ge-scale, gene al popula ion s udies
ha e examined changes in pu chases o sho -ac ing β
2−
-agonis s ol-
lowing ecen s ess ul li e e en s among adul s wi h diagnosed as hma.
2. Me hod
2.1. S udy design and pa icipan s
The Heal h and Social Suppo S udy, a longi udinal coho s udy, is
based on a ep esen a i e sample o he Finnish popula ion in he age
g oups: 20–24, 30–34, 40–44, and 50–54 yea s a baseline [24]. The
baseline pos al su ey was conduc ed in 1998, and a o al o 25,901
esponden s e u ned he ques ionnai e. O hem, 19,629 esponden s
(80% o hose eligible) pa icipa ed in a ollow-up su ey 5 yea s la e
in 2003 and 18,900 (96%) consen ed o he use o hei eco ded heal h
in o ma ion om he Finnish na ional egis e s. Fo his s udy, we se-
lec ed all pa icipan s o he ollow-up su ey wi h as hma (n= 1102,
73% women) a he beginning o he 7-yea obse a ion window om
Janua y 1, 2000 onwa d who had p o ided in o ma ion on he occu -
ence o new li e e en s wi hin he p eceding six mon hs in he ollow-
up su ey. (Fig. 1) The s udy was app o ed by he Tu ku Uni e si y
Hospi al E hics Commi ee. All pa icipan s signed an in o med consen
o m.
2.2. Pa icipan s wi h as hma
We used he unified pe sonal iden ifica ion code sys em, co e ing
all Finnish ci izens, o link and ob ain eco ds om h ee adminis a i e
and comp ehensi e Finnish na ional heal h egis e s o iden i y in-
di iduals wi h as hma and hei pu chases o p esc ibed as hma medi-
ca ions.
The iden ifica ion o a pa icipan ha ing as hma a he beginning o
he ollow-up was based on he clinical diagnosis o he ea ing phy-
sician in he eco ds o he D ug Reimbu semen Regis e o he Social
Insu ance Ins i u ion (SII) o Finland [25] and/o he Hospi al Dis-
cha ge Regis e o he Na ional Ins i u e o Heal h and Wel a e. We
used he D ug Reimbu semen Regis e o he SII o Finland con aining
in o ma ion on pe sons en i led o special eimbu semen o ce ain
ch onic diseases, such as as hma. Pa ien s who apply o special e-
imbu semen mus a ach a de ailed medical ce ifica e p epa ed by he
ea ing physician, who also p o ides da a o confi m he diagnosis. The
applica ion is hen e iewed by a physician in he SII o de e mine
whe he he uni o mly defined equi emen s o he disease a e me .
F om his egis e , pa icipan s we e defined as as hma cases i hey
we e o he fi s ime eco ded in he Cen al D ug Regis e as eligible
o as hma ea men be o e he s a o he ollow-up in Janua y 1,
2000. Mo eo e , we used p esc ip ion da a o assess he beginnings o
medical ea men o as hma. In Finland, he Na ional Social Insu ance
Scheme a he SII p o ides basic eimbu semen o all filled ou pa ien
p esc ip ions ha a e eco ded in he D ug P esc ip ion Regis e ac-
co ding o he Wo ld Heal h O ganiza ion's Ana omical The apeu ic
Chemical (ATC) Classifica ion. The da e o pu chase is also eco ded.
We iden ified all pa icipan s wi h wo o mo e p esc ip ions o d ugs
o obs uc i e ai way diseases (ATC code R03) in 1998 and 1999 ( he
wo yea s p eceding he beginning o he ollow-up) by using he day o
he fi s pu chase as an indica o o p e alen as hma. Finally, we ob-
ained da a om he Hospi al Discha ge Regis e o he Na ional In-
s i u e o Heal h and Wel a e, which includes eco ds on all inpa ien
hospi al admissions [26]. This egis e is comp ised o coun ywide
in o ma ion on i ually all hospi aliza ions. All pa icipan s discha ged
om hospi als wi h he main diagnosis ICD-10 J45 (as hma) wi hin he
wo yea s p eceding he beginning o he ollow-up we e also defined as
as hma cases.
2.3. Assessmen o filled p esc ip ion o as hma medica ion du ing he 7-
yea ollow-up
In Finland, inhaled as hma medica ions a e only a ailable by p e-
sc ip ion. The Na ional Heal h Insu ance Scheme, un by he SII o
Finland, p o ides p esc ip ion d ug co e age o all (~5.5 million)
communi y-dwelling esiden s o Finland. All eimbu sed p esc ip ions
a e egis e ed in he D ug P esc ip ion Regis e managed by he SII
[25]. Fo each d ug, he dispensing da e and he Wo ld Heal h O ga-
niza ion Ana omical The apeu ic Chemical (ATC) code a e eco ded.
We de i ed he da e and he ATC classifica ion code o pu chases o
inhaled β
2
-agonis s and co icos e oids du ing a se en-yea obse a ion
pe iod co e ing he yea s 2000 o 2006 om he D ug P esc ip ion
Regis e . Fo obse a ion, we de e mined he numbe o pu chases o
inhaled sho -ac ing β
2
- agonis s (ATC R03AC02, R03AC03 R03AC04)
and combina ions o sho -ac ing β
2
-agonis s wi h an icholine gics
(ATC R03AK03, R03AK04).
Because he need o sho -ac ing β
2
-agonis s may depend on he
simul aneous usage o inhaled an i-inflamma o y medica ion, we de-
e mined he numbe o pu chases o inhaled co icos e oids (ATC
R03BA01, R03BA02, R03BA05) and long-ac ing β
2
-agonis s (ATC
R03AC12, R03AC13) in non-combina ion inhale s and fixed dose
combina ion inhale s o inhaled co icos e oids and long-ac ing β
2
-
agonis s (ATC R03AK06, R03AK07). The numbe o pu chases in e e y
yea o obse a ion was handled as a ca ego ical ac o in he model.
2.4. Recen s ess ul li e e en s
We measu ed he occu ence o ecen s ess ul li e e en s in he
ollow-up su ey conduc ed in 2003 by using 19 li e e en s om a lis
o 21 e en s [27,28]. The excluded e en s (i.e., “illness causing wo k
disabili y o o e 21 days”and “disabili y e i emen ”) migh ha e been
a consequence o as hma exace ba ion (see Appendix 1). Fo he iming
o each e en , he ques ionnai e included ou esponse al e na i es
(ne e , wi hin he p e ious 6 mon hs, wi hin he p e ious 5 y s.
and > 5 y s. ago), and he esponden s we e ins uc ed o selec only
one o hem. The ocus o his s udy is in ecen e en s, hose ha had
occu ed du ing he p e ious six mon hs. We assessed he le el o
R. Lie zén e al. Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
47
exposu e o ecen li e e en s o each indi idual by calcula ing a cu-
mula i e mean sum sco e weigh ed by he a e age se e i y o he e en
( o weigh s, see Vah e a e al. [28]). The cumula i e se e i y a ings
anged om 2.74 o 24.64. Those esponden s wi h a 0 (ze o) sco e
we e defined as ha ing no exposu e. Those who had been exposed we e
di ided in o wo g oups using he median o he cumula i e se e i y
sco e as he cu -offpoin (low exposu e < 5, high exposu e ≥5).
As shown in Appendix 1, only a small p opo ion o pa icipan s had
epo ed he co esponding e en in he 1998 su ey (e.g. 14% o hose
who we e ic ims o iolence in 2003 epo ed o ha e been ic ims in
1998 also). E en s wi h highes ecu ence - b eakup o long- e m
iendship (43%), se e e financial difficul ies (40%) and dea h o a
close ela i e (35%) - we e a ed much less se e e and, hus, had a
lowe weigh in he cumula i e mean sum sco e.
2.5. Baseline cha ac e is ics
Baseline cha ac e is ics, measu ed in he baseline su ey in 1998
be o e exposu e o he li e e en s included socio-demog aphic a iables
–sex, age g oup, ma i al s a us, and le el o educa ion –beha iou -
ela ed heal h isks, sensi i i y o s ess and dep ession. The beha iou -
ela ed heal h isks we e smoking (ne e /ex/cu en ), high alcohol
in ake (≥175 g o alcohol o women and ≥263 g o alcohol o men
pe week) [29], obesi y (Body Mass Index (BMI) ≥30 kg/m
2
) and
physical inac i i y ( he Me abolic Equi alen Task index < 2 MET-
hou s/day) [30]. Indi idual diffe ences in sensi i i y o s ess we e
measu ed by gene al eelings o s ess ulness in daily li e [28,31]. The
mean sco es o he scale we e di ided wi h e iles, wi h he highes
hi d used as an indica o o sensi i i y o s ess. Dep ession, a po en ial
media o be ween a s ess ul li e e en and as hma, was assessed using
he Beck Dep ession In en o y (sum sco e > 18) [32] and he D ug
P esc ip ion Regis e [≥1 an idep essan (ATC-code N06A) pu chases
in 1998]. Pa icipan s showing dep ession in any o hese measu e-
men s we e classified as cases o p e-exis ing dep ession.
2.6. S a is ical analysis
The associa ion be ween backg ound a iables (demog aphics,
heal h- ela ed ac o s, and psychological ac o s) (measu ed in 1998)
and ecen s ess ul li e-e en exposu e (measu ed in 2003) ca ego ies
we e s udied using he Pea son's chi-squa ed es s.
Fo he analyses, we di ided he se en-yea ollow-up ime in o
h ee pe iods in ela ion o he iming o he li e-e en exposu e: ‘be-
o e’(i.e., yea −3 and −2, 2000–2001), ‘du ing’(i.e. yea −1 and 0,
2002–2003) and ‘a e ’(i.e. yea +1 o +3, 2004–2006). We applied a
epea ed-measu es Poisson eg ession analysis wi h he gene alized
es ima ing equa ion (GEE) me hod and au o eg essi e co ela ion
s uc u e [33]. The GEE akes in o accoun he co ela ion o annual
medica ion pu chases wi hin pe sons, and is no e y sensi i e o
missing cases a epea ed measu emen s. We fi s examined he pu -
chases o inhaled sho -ac ing b onchodila o s by baseline cha ac e -
is ic by calcula ing he mean a es o pu chases o e he se en-yea
obse a ion pe iod o each cha ac e is ic.
We used con as s o es ima e he a e a ios and hei 95% con-
fidence in e als (95% CI) du ing and a e he exposu e compa ed wi h
he pe iod be o e he exposu e - wi hin each exposu e g oup using
Fig. 1. Sample selec ion o subjec s wi h as hma exposed o e-
cen s ess ul li e e en s and linked o medica ion eco ds
2000–2006, he Heal h and Social Suppo S udy in Finland,
1998–2006. The iden ifica ion o a esponden s ha ing as hma
was based on na ional egis e s.
R. Lie zén e al. Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
48
models including he in e ac ion e m “s ess ul li e-e en ex-
posu e*pe iod”. The a e a ios o pu chases o inhaled sho -ac ing β
2
-
agonis s we e calcula ed in he h ee ime pe iods a ound he li e-e en
exposu e. The analyses we e adjus ed o all baseline cha ac e is ics. An
addi ional adjus men was made o inhaled an i-inflamma o y medi-
ca ion. In o de o examine whe he changes in pu chases o inhaled
sho -ac ing β
2
-agonis s ollowing exposu e o li e e en s a ied be-
ween he subg oups (e.g. by sex, age g oup, educa ion, beha iou - e-
la ed heal h isks, dep ession o sensi i i y o s ess) we calcula ed he
a e a ios o each subg oup by using he same model. All es s we e 2-
ailed.
All analyses we e pe o med using he SAS En e p ise Guide 6.100
(6.100.0.2870) s a is ical so wa e (SAS Ins i u e Inc., Ca y, NC, USA,
2013).
3. Resul s
3.1. Cha ac e is ics o he s udy popula ion
The sample included 296 (27%) men and 806 (73%) women wi h
p e alen as hma a baseline. O hese indi iduals, 367 (33%) epo ed
an occu ence o new ecen s ess ul li e e en s in he ollow-up su ey.
The e en s epo ed mos o en we e: ‘majo inc ease in ma i al p o-
blems’(n= 83), ‘se e e financial difficul ies’(n= 77) o /and ‘dea h o
ano he close ela i e’(n= 60). (Appendix 1). Table 1 shows he as-
socia ions be ween he se e i y o li e e en s exposed (no/low/high)
and he cha ac e is ics o pa icipan s; a young age, cu en o ex
smoking, and physical ac i i y we e associa ed wi h high li e-e en
exposu e.
3.2. Pu chases o inhaled sho -ac ing β
2
-agonis s
Du ing he se en-yea obse a ion pe iod, 5955 pu chases o in-
haled sho -ac ing β
2
-agonis s we e eco ded o he pa icipan s. A
high annual pu chase a e was obse ed among smoke s, as well as
hose wi h dep ession, obesi y and hose wi h a basic le el o educa ion
(Table 1).
3.3. Inhaled sho -ac ing β
2
-agonis s and ecen s ess ul li e e en s
Table 2 shows he a e a ios o pu chases o inhaled sho -ac ing
β
2
-agonis s in he ime pe iods du ing (yea −1 and 0) and a e (yea
Table 1
Cha ac e is ics o he pa icipan a baseline in 1998 by ecen s ess ul li e-e en exposu e le els in 2003 and annual mean a es o pu chases o inhaled sho -ac ing β
2
-agonis s du ing
2000–2006, he heal h and social suppo s udy in Finland, 1998–2006.
Le el o li e-e en exposu e
All pa icipan s No Low High Pu chases 2000–2006
No. % No. % No. % No. % PValue
a
Mean a e
b
PValue
c
To al 1102 100 73 66 205 19 162 15
Sex 0.54 0.96
Men 296 27 200 27 58 28 38 23 77.6
Women 806 73 535 73 147 72 124 77 77.1
Age g oup < 0.001 0.20
20–24 242 22 135 18 57 28 50 31 64.1
30–34 239 22 145 20 53 26 41 25 79.2
40–44 252 23 171 23 48 23 33 20 71.5
50–54 369 33 284 39 47 23 38 24 88.5
Occupa ional educa ion 0.070 0.035
Uni e si y 177 16 129 18 29 14 19 12 59.7
College 335 31 211 29 74 37 50 31 69.4
Voca ional school 237 22 158 22 34 17 45 28 75.0
Basic 336 31 225 31 64 32 47 29 94.5
Ma i al s a us 0.015 0.24
Single/di o ced/widowed 326 30 199 27 65 32 62 38 85.4
Ma ied/cohabi ing 775 70 535 73 140 68 62 62 73.5
Smoking 0.003 < 0.001
Ne e -smoke 449 45 328 49 72 39 49 32 60.4
Ex-smoke 296 29 186 28 56 30 54 36 66.1
Cu en smoke 266 26 161 24 57 31 48 32 124.8
Physical inac i i y 0.008 0.41
No 826 75 532 73 171 83 123 76 84.2
Yes 270 25 198 27 34 17 38 24 74.2
Obesi y (BMI ≥30)
d
0.53 0.007
No 921 84 609 83 173 84 139 87 68.8
Yes 175 16 122 17 32 16 21 13 124.0
High alcohol in ake
e
0.43 0.72
No 1046 95 701 96 193 94 152 94 77.0
Yes 53 5 31 4 12 6 10 6 84.1
Dep ession 0.36 0.036
No 981 89 659 90 183 89 139 86 73.5
Yes 121 11 76 10 22 11 23 14 107.4
Sensi i i y o s ess 0.066 0.22
No 720 66 496 68 127 62 97 60 72.4
Yes 374 34 232 32 77 38 65 40 86.3
Abb e ia ions: BMI, body mass index.
a
P alue o diffe ence be ween he exposu e g oups (Pea son's chi-squa ed es ).
b
Annual a e age pu chases o inhaled sho -ac ing β
2
-agonis s pe 100 pe son yea s de i ed om Poisson eg ession gene alized es ima ing equa ion (GEE) analysis o co a ia e.
c
P alue om Poisson eg ession GEE analysis o mean a es.
d
BMI was calcula ed as weigh (kg)/heigh (m
2
).
e
High alcohol in ake e e s o consump ion o > 175 g/week o women and > 263 g/week o men.
R. Lie zén e al. Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
49
+1 o +3), compa ed wi h he pe iod be o e (yea −3 and −2),
exposu e o s ess ul li e e en s adjus ed o he baseline cha ac e is ics
(demog aphics, heal h isk beha iou s, dep ession and sensi i i y o
s ess) and, addi ionally, o inhaled an i-inflamma o y medica ion.
Among hose wi h no exposu e o ecen s ess ul li e e en s, he a e o
pu chases did no significan ly a y be ween he ime pe iods. Among
pa icipan s who had encoun e ed highly s ess ul ecen li e e en s, he
pu chases o inhaled sho -ac ing β
2
-agonis s inc eased by 46% du ing
he li e e en exposu e and 52% a e he exposu e compa ed wi h he
p e-exposu e le els. This a e a io emained unchanged a e aking
in o accoun adjus men s o all baseline cha ac e is ics. Addi ionally,
adjus men o inhaled an i-inflamma o y medica ion only sligh ly a -
enua ed he associa ion (p< 0.06). Among pa icipan s who had low
exposu e o ecen s ess ul e en s, no inc ease be ween pe iods exis ed.
Resul s om he subg oup analyses a e shown in Table 3. Compa ed
wi h he p e-exposu e le el, he a e o pu chases o inhaled sho -
ac ing β
2
-agonis s inc eased in all subg oups o pa icipan s du ing and
a e high li e-e en exposu e. One excep ion was ela ed o age; a e
exposu e, he younges age g oup showed no inc ease in he a e o
pu chases o inhaled sho -ac ing β
2
-agonis s. In e es ingly, among
dep essi e pa icipan s (n= 121), he pos -exposu e inc ease was ex-
cep ionally high, 4.4- old compa ed o he p e-exposu e le el.
Appendix 1 p esen s how many o he 367 pa icipan s epo ing a
ecen e en a he ollow-up su ey in 2003 epo ed he co esponding
e en also in he 1998 su ey. As can be seen, o he mos se e e e en s
ecu ence was non-exis ing o a e (e.g. only 14% o hose who we e
ic ims o iolence in 2003 epo ed o ha e been ic ims in p eceding
six mon hs in 1998 also). Highes ecu ence was ound o ‘b eakup o
long- e m iendship’(43%), ‘se e e financial difficul ies’(40%) and
‘dea h o a close ela i e’(35%), e en s a ed much less se e e.
4. Discussion
In his popula ion-based 7-yea ollow-up s udy on adul s wi h
as hma, eco ds o filled p esc ip ions o inhaled sho -ac ing β
2
-ago-
nis s be o e, du ing and a e exposu e o s ess ul li e-e en s we e used
as an indica o o as hma symp oms. Among hose who had en-
coun e ed ecen highly s ess ul li e e en s, he a e o inhaled sho -
ac ing β
2
-agonis s pu chases was 1.5 imes highe a e he li e e en
compa ed o he p e-e en le els. No such inc ease was obse ed among
hose wi h no o low exposu e o s ess ul li e e en s. These findings
suppo he hypo hesis ha psychosocial s ess may exace ba e as hma
symp oms in wo king-aged adul s.
Ou esul s a e consis en wi h he biopsychosocial model o s ess,
which sugges s ha s ess ul li e e en s may al e he psychological,
immunological and endoc ine sys ems in ways ha lead o he ex-
ace ba ion o as hma [9,34]. Mo eo e , he esul s o his s udy a e in
line wi h he ew s udies which in es iga ed he associa ion be ween
s ess ul li e e en s and he exace ba ion o as hma symp oms in adul s.
A case-con ol s udy by Kolbe e al. [19] ound ha li e e en s we e
epo ed mo e o en among pa ien s admi ed o hospi al wi h acu e
as hma compa ed o a con ol g oup o non-hospi alized as hma ics. In
a p ospec i e popula ion based s udy by Wainw igh e al. [18] li e
e en s expe ienced in adul hood we e associa ed wi h inc eased a es o
as hma- ela ed hospi al admissions. Al hough ea lie s udies p o ide
impo an insigh on as hma exace ba ion, hospi al admissions o
as hma a e conside ed e y a e [21], ep esen ing only he ip o an
icebe g o as hma mo bidi y. In his s udy, he finding ha he u ili-
za ion o inhaled sho -ac ing β
2
-agonis s a ies depending on he ex-
en o which as hma ic adul s a e exposed o li e s ess has no been
p e iously epo ed.
S ess may ha e an impac on indi iduals' li e managemen and hus
affec s as hma sel -ca e and adhe ence o ea men [34], and inc eases
he isk o inapp op ia e use o as hma medica ions [35]. A ecen s udy
ound ha many pa ien s pe cei ed s ess was an impo an de e mi-
nan o uncon olled as hma [36]. One indica o o such a de elopmen
is high use o sho -ac ing b onchodila o s wi h low use o inhaled
s e oids [35]. As we we e able o con ol in he analysis simul aneous
usage o an i-inflamma o y medica ion, poo sel -ca e o as hma is an
unlikely explana ion o ou ime-dependen findings.
The s eng hs o his s udy a e i s la ge sample size and a s udy
design ha allowed he de e mina ion o empo a y o de be ween
exposu e o ecen s ess ul li e e en s and pu chases o as hma medi-
ca ion. The numbe o filled p esc ip ions o inhaled sho -ac ing β
2
-
agonis s and he measu emen o as hma we e based on na ional heal h
egis e s. In Finland, he alidi y o he na ional egis e s has been
ound o be high [37], easonably accu a e, and highly eliable o
epidemiological s udy pu poses [38]. We we e also able o con ol a
numbe o socio-demog aphic elemen s and e iological ac o s o
as hma.
A limi a ion o he s udy is ha filled p esc ip ions do no equa e o
ac ual medica ion u iliza ion. Because we did no ha e in o ma ion on
o al amoun s o d ugs pe p esc ip ion o ecommended doses o he
as hma ea men , we we e no able o use mo e fine-g ained measu es
such as inc eased daily doses o inhaled sho -ac ing b onchodila o s
use. Howe e , sho -ac ing be a-agonis p esc ip ion fills can be used as
a ma ke o as hma mo bidi y [39]. A baseline, he esponse a e was
ela i ely low (40%), and he e may ha e been diffe ences be ween
esponden s and non- esponden s ega ding he equency o as hma
and ecen s ess ul li e e en s, al hough no majo heal h- ela ed se-
lec ion has been de ec ed in a non- esponse analysis [24]. How-
e e , > 80% o he baseline esponden s pa icipa ed in he ollow-up
su ey, and p ac ically all o hem (96%) consen ed o he linking o
da a om na ional heal h egis e s. Thus, i is unlikely ha he long-
i udinal associa ion be ween ecen s ess ul li e e en s and as hma
exace ba ion would be biased due o low pa icipa ion a baseline.
Addi ionally, due o he applied su ey me hodology, we did no ha e
any addi ional in o ma ion o he epo ed li e e en s and hei s ess-
ulness. Howe e , by using a e age se e i y a ing ins ead o indi idual
pe cep ion o he se e i y o he e en ou measu e was no con ounded
by he consequences o he e en o he indi idual.
Table 2
Ra e Ra ios o Pu chases o Inhaled Sho -Ac ing β
2
-Agonis s Compa ing Diffe en Time
Pe iods Acco ding o Li e-E en exposu e, The Heal h and Social Suppo S udy in
Finland, 1998–2006.
Adjus men le el o s ess ul li e-
e en exposu e
Time in ela ion o li e-e en exposu e
Du ing
a
s be o e
c
A e
b
s be o e
c
RR
d
95% CI RR
d
95% CI
Unadjus ed
No exposu e 0.98 0.88, 1.09 0.94 0.82, 1.07
Low exposu e 0.84 0.75, 0.95 0.89 0.72, 1.10
High exposu e 1.46 1.12, 1.90 1.52 1.06, 2.16
Baseline adjus ed
e
No exposu e 0.99 0.88, 1.11 0.95 0.83, 1.09
Low exposu e 0.84 0.74, 0.96 0.81 0.66, 0.99
High exposu e 1.47 1.12, 1.92 1.50 1.05, 2.13
Addi ionally adjus ed
No exposu e 0.97 0.85, 1.10 0.92 0.79, 1.07
Low exposu e 0.79 0.69, 0.91 0.74 0.62, 0.88
High exposu e 1.43 1.09, 1.87 1.43 0.99, 2.07
Abb e ia ions: CI, confidence in e al; RR, a e a io.
a
Du ing e e s o yea −1 and 0 in ela ion o he iming o he li e-e en exposu e.
b
A e e e s o yea +1 o +3 in ela ion o he iming o he li e-e en exposu e.
c
Be o e e e s o yea −3 and −2 in ela ion o he iming o he li e-e en exposu e.
d
Ra e a ios (RR) and hei 95% confidence limi s (CI) de i ed om Poisson eg ession
gene alized es ima ing equa ion analysis o ime pe iods.
e
Adjus ed o sex, age, educa ion, ma i al s a us, smoking, seden a y li es yle, obesi y,
high alcohol in ake, dep ession and sensi i i y o s ess.
Addi ionally adjus ed o inhaled co icos e oids and long-ac ing β
2
-agonis s in
combina ion o sepa a e as ime dependen a iable.
R. Lie zén e al. Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
50

5. Conclusions
To he bes o ou knowledge, his is he fi s s udy o in es iga e he
longi udinal associa ions be ween ecen s ess ul li e e en s and pu -
chase le els in a popula ion sample o adul s wi h as hma a middle age.
Ou finding ha exposu e o highly-s ess ul li e e en s is associa ed
wi h an inc ease in inhaled sho -ac ing β
2
-agonis s pu chases sugges a
wo sening o as hma symp oms in he a e ma h o s ess ul expe i-
ences. These esul s highligh he po en ial impo ance o aking in o
accoun psychosocial s ess in guiding as hma sel -ca e.
Supplemen a y da a o his a icle can be ound online a h p://dx.
doi.o g/10.1016/j.jpsycho es.2017.07.003.
Compe ing In e es S a emen
All au ho s ha e comple ed he Unified Compe ing In e es o m a
h p://www.icmje.o g/coi_disclosu e.pd and decla e ha (1) D .
Ki imäki ecei ed suppo om No dFo sk, he Medical Resea ch
Council, and he Economic and Social Resea ch Council, du ing he
conduc o he s udy; (2) au ho s ha e no ela ionships wi h companies
o o he compe ing in e es s in he pas h ee yea s ha could be pe -
cei ed o cons i u e a conflic o in e es ; (3) spouses, pa ne s, o
child en o au ho s ha e no financial ela ionships ha may be ele an
o he submi ed wo k; and (4) au ho s ha e no non-financial in e es s
ha may be ele an o he submi ed wo k.
Acknowledgemen s
The Heal h and Social Suppo S udy was suppo ed by he Social
Insu ance Ins i u ion (SII) o Finland; he Ida Mon in Founda ion,
Finland (g an o RL); he Resea ch Founda ion o he Pulmona y
Diseases in Finland (g an o RL); he Alle gy Founda ion in Finland
(g an o RL); No dFo sk, he No dic P og amme on Heal h and Wel a e
(g an 75021 o MKi); he Medical Resea ch Council, Uni ed Kingdom
(g an K013351 o MKi); and he Economic and Social Resea ch
Council, Uni ed Kingdom (g an o MKi).
Re e ences
[1] E.D. Ba eman, S.S. Hu d, P.J. Ba nes, e al., Global s a egy o as hma managemen
Table 3
Ra e Ra ios o Pu chases o Inhaled Sho -Ac ing β
2
-Agonis s Compa ing Diffe en Time Pe iods by Recen S ess ul Li e-E en Exposu e Le els Acco ding o Baseline Cha ac e is ics, The
Heal h and Social Suppo S udy in Finland, 1998–2006.
No s ess ul li e-e en exposu e Low s ess ul li e-e en exposu e High s ess ul li e-e en exposu e
Time in ela ion o li e-e en exposu e Time in ela ion o li e-e en exposu e Time in ela ion o li e-e en exposu e
Du ing
a
s be o e
b
A e
c
s be o e
b
Du ing
a
s be o e
b
A e
c
s be o e
b
Du ing
a
s be o e
b
A e
c
s be o e
b
Cha ac e is ic RR
d
95% CI RR
d
95% CI RR
d
95% CI RR
d
95% CI RR
d
95% CI RR
d
95% CI
Sex
Men 0.89 0.75, 1.06 0.93 0.77, 1.13 0.94 0.70. 1.26 1.02 0.67, 1.57 1.22 0.70, 2.10 1.38 0.59, 3.23
Women 1.00 0.88, 1.15 0.94 0.79, 1.11 0.82 0.72, 1.92 0.85 0.67, 1.08 1.59 1.21, 2.09 1.59 1.17, 2.15
Age g oup
20–24 0.84 0.62, 1.13 0.85 0.64, 1.14 0.90 0.63, 1.29 1.06 0.67, 1.67 1.20 0.84, 1.71 0.64 0.35, 1.15
30–34 1.05 0.83, 1.34 1.20 0.87, 1.66 0.91 0.72, 1.16 0.92 0.59, 1.44 1.38 0.75, 2.55 2.29 1.13, 4.64
40–44 0.98 0.76, 1.27 0.94 0.69, 1.28 0.77 0.63, 0.95 0.89 0.68, 1.15 1.66 0.96, 2.86 1.96 1.22, 3.14
50–54 0.99 0.86, 1.15 0.87 0.73, 1.03 0.73 0.59, 0.89 0.70 0.44, 1.11 1.76 1.14, 2.71 2.31 1.45, 3.70
Educa ion
Uni e si y 0.97 0.69, 1.37 1.01 0.70, 1.47 0.80 0.64, 1.00 0.79 0.48, 1.32 1.21 0.55, 2.66 1.41 0.60, 3.34
College 0.91 0.74, 1.11 0.84 0.64, 1.11 0.90 0.68, 1.18 1.04 0.71, 1.52 1.57 0.96, 2.56 1.81 1.04, 3.14
Voca ional school 0.98 0.79, 1.21 0.94 0.73, 1.21 0.83 0.65, 1.07 0.49 0.34, 0.72 1.13 0.75, 1.70 1.03 0.51, 2.06
Basic 1.04 0.87, 1.25 0.99 0.80, 1.22 0.85 0.71, 1.02 1.04 0.73, 1.46 1.89 1.28, 2.78 2.03 1.25, 3.30
Ma i al s a us
Single/di o ced/widowed 0.96 0.79, 1.16 0.96 0.77, 1.21 0.91 0.75, 1.10 0.89 0.64, 1.26 1.31 0.95, 1.83 1.15 0.76, 1.74
Ma ied/cohabi ing 0.98 0.86, 1.12 0.92 0.78, 1.09 0.81 0.69, 0.94 0.87 0.67, 1.14 1.57 1.06, 2.31 1.80 1.09, 2.98
Smoking
Ne e -smoke 0.92 0.78, 1.07 0.88 0.74, 1.06 0.80 0.61, 1.05 0.89 0.60, 1.30 1.39 0.95, 2.02 1.40 0.97, 2.01
Ex-smoke 1.20 0.95, 1.51 1.16 0.91, 1.49 0.89 0.74, 1.08 0.77 0.55, 1.08 1.23 0.71, 2.16 1.34 0.60, 2.95
Cu en smoke 0.96 0.79, 1.16 0.96 0.75, 1.23 0.83 0.69, 1.01 0.80 0.57, 1.12 1.65 1.04, 2.62 1.62 0.84, 3.12
Physical inac i i y
No 1.00 0.88, 1.15 0.95 0.81, 1.11 0.81 0.71, 0.93 0.92 0.73, 1.17 1.35 1.04, 1.75 1.28 0.86, 1.90
Yes 0.90 0.77, 1.07 0.90 0.72, 1.13 0.99 0.73, 1.34 0.77 0.51, 1.18 1.97 0.84, 4.61 2.64 1.17, 5.95
Obesi y (BMI ≥30)
e
No 0.98 0.86, 1.12 0.94 0.80, 1.09 0.85 0.74, 0.98 0.87 0.69, 1.10 1.39 1.02, 1.89 1.39 0.93, 2.09
Yes 0.96 0.81, 1.14 0.93 0.74, 1.18 0.76 0.62, 0.92 0.94 0.62, 1.41 1.82 1.28, 2.57 2.19 1.39, 3.47
High alcohol in ake
No 0.97 0.87, 1.08 0.92 0.80, 1.06 0.82 0.72, 0.93 0.87 0.69, 1.10 1.44 1.10, 1.89 1.47 1.02, 2.12
Yes 1.18 0.74, 1.90 1.46 0.83, 2.57 1.07 0.89, 1.29 1.06 0.73, 1.53 2.23 0.66, 7.53 3.76 0.76, 18.50
Dep ession
No 0.99 0.88, 1.12 0.96 0.83, 1.11 0.84 0.73, 0.96 0.93 0.73, 1.18 1.34 1.02, 1.75 1.30 0.90, 1.88
Yes 0.90 0.73, 1.12 0.84 0.64, 1.11 0.86 0.66, 1.13 0.70 0.46, 1.07 2.93 1.47, 5.85 4.36 1.75, 10.84
Sensi i i y o s ess
No 0.97 0.84, 1.11 0.93 0.80, 1.09 0.77 0.67, 0.90 0.83 0.64, 1.07 1.36 0.97, 1.92 1.29 0.84, 1.97
Yes 0.99 0.83, 1.18 0.92 0.73, 1.15 0.97 0.78, 1.19 0.98 0.67, 1.42 1.56 1,00, 2.41 1.77 0.97, 3.25
Abb e ia ions: BMI, body mass index; CI, confidence in e al; RR, a e a io.
a
Du ing e e s o yea −1 and 0 in ela ion o he iming o he li e-e en exposu e.
b
A e e e s o yea +1 o +3 in ela ion o he iming o he li e-e en exposu e.
c
Be o e e e s o yea −3 and −2 in ela ion o he iming o he li e-e en exposu e.
d
Ra e a ios (RR) and hei 95% confidence limi s (CI) de i ed om Poisson eg ession gene alized es ima ing equa ion analysis o ime pe iods.
e
BMI was calcula ed as weigh (kg)/heigh (m
2
).
High alcohol in ake e e s o consump ion o > 175 g/week o women and > 263 g/week o men.
R. Lie zén e al. Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
51
and p e en ion: GINA execu i e summa y, Eu . Respi . J. 31 (2008) 143–178.
[2] D.J. Jackson, A. Sykes, P. Mallia, e al., As hma exace ba ions: o igin, effec , and
p e en ion, J. Alle gy Clin. Immunol. 128 (2011) 1165–1174.
[3] To T, S. S anoje ic, G. Moo es, e al., Global as hma p e alence in adul s: findings
om he c oss-sec ional wo ld heal h su ey, BMC Public Heal h 12 (2012) (204-
2458-12-204).
[4] GBD, Disease and inju y incidence and p e alence collabo a o s. Global, egional,
and na ional incidence, p e alence, and yea s li ed wi h disabili y o 310 diseases
and inju ies, 1990–2015: a sys ema ic analysis o he Global Bu den o Disease
S udy 2015, Lance 2016 (388) (2015) 1545–1602.
[5] W.W. Busse, R.F. Lemanske J ., J.E. Ge n, Role o i al espi a o y in ec ions in
as hma and as hma exace ba ions, Lance 376 (2010) 826–834.
[6] M.K. Ve non, Wha do we know abou as hma igge s? A e iew o he li e a u e, J.
As hma 49 (2012) 991–998.
[7] J. Douwes, C. B ooks, N. Pea ce, As hma ne osa: old concep , new insigh s, Eu .
Respi . J. 37 (2011) 986–990.
[8] M.A. Yonas, N.E. Lange, J.C. Celedon, Psychosocial s ess and as hma mo bidi y,
Cu . Opin. Alle gy Clin. Immunol. 12 (2012) 202–210.
[9] R.J. W igh , R.T. Cohen, S. Cohen, The impac o s ess on he de elopmen and
exp ession o a opy, Cu . Opin. Alle gy Clin. Immunol. 5 (2005) 23–29.
[10] N.D. Da e, L. Xiang, K.E. Rehm, e al., S ess and alle gic diseases, Immunol. Alle gy
Clin. N. Am. 31 (2011) 55–68.
[11] A.F. T ueba, T. Ri z, S ess, as hma, and espi a o y in ec ions: pa hways in ol ing
ai way immunology and mic obial endoc inology, B ain Beha . Immun. 29 (2013)
11–27.
[12] M. Kilpeläinen, M. Kosken uo, H. Helenius, e al., S ess ul li e e en s p omo e he
mani es a ion o as hma and a opic diseases, Clin. Exp. Alle gy 32 (2002) 256–263.
[13] A. Loe b oks, C.J. Ap elbache , J.F. Thaye , e al., Neu o icism, ex a e sion,
s ess ul li e e en s and as hma: a coho s udy o middle-aged adul s, Alle gy 64
(2009) 1444–1450.
[14] R. Lie zen, P. Vi anen, M. Ki imaki, e al., S ess ul li e e en s and he onse o
as hma, Eu . Respi . J. 37 (2011) 1360–1365.
[15] A.M. Renzaho, B. Houng, J. Old oyd, e al., S ess ul li e e en s and he onse o
ch onic diseases among Aus alian adul s: findings om a longi udinal su ey, Eu .
J. Pub. Heal h 24 (2014) 57–62.
[16] A. Loe b oks, H. Ding, W. Han, e al., Wo k s ess, amily s ess and as hma: a c oss-
sec ional s udy among women in China, In . A ch. Occup. En i on. Heal h 90
(2017) 349–356.
[17] A.J. Ap e , L.A. Ga cia, R.C. Boyd, e al., Exposu e o communi y iolence is asso-
cia ed wi h as hma hospi aliza ions and eme gency depa men isi s, J. Alle gy
Clin. Immunol. 126 (2010) 552–557.
[18] N.W. Wainw igh , P.G. Su ees, N.J. Wa eham, e al., Psychosocial ac o s and in-
ciden as hma hospi al admissions in he EPIC-No olk coho s udy, Alle gy 62
(2007) 554–560.
[19] J. Kolbe, W. Fe gusson, M. Vamos, e al., Case-con ol s udy o se e e li e h ea-
ening as hma (SLTA) in adul s: psychological ac o s, Tho ax 57 (2002) 317–322.
[20] K. Heikkila, I.E. Madsen, S.T. Nybe g, e al., Job s ain and he isk o se e e as hma
exace ba ions: a me a-analysis o indi idual-pa icipan da a om 100 000
Eu opean men and women, Alle gy 69 (2014) 775–783.
[21] P. Kauppi, M. Linna, J. Ma ikainen, e al., Follow-up o he Finnish As hma
P og amme 2000–2010: educ ion o hospi al bu den needs isk g oup e hinking,
Tho ax 68 (2013) 292–293.
[22] T. Haah ela, T. Klaukka, K. Koskela, e al., As hma p og amme in Finland: a com-
muni y p oblem needs communi y solu ions, Tho ax 56 (2001) 806–814.
[23] Global S a egy o As hma Managemen and P e en ion. Global Ini ia i e o
As hma (GINA). www.ginas hma.o g. (Las Upda ed: 2017. Assessed May 1, 2017).
[24] K. Ko keila, S. Suominen, J. Ah enainen, e al., Non- esponse and ela ed ac o s in
a na ion-wide heal h su ey, Eu . J. Epidemiol. 17 (2001) 991–999.
[25] The Social Insu ance Ins i u ion o Finland - Kela. S a is ics on eimbu semen en-
i lemen s in espec o medicines. h p://www.kela.fi/web/en/s a is ics-by- opic_
s a is ics-on- eimbu semen -en i lemen s-in- espec -o -medicines. (Published
Oc obe 22, 2012. Da e las upda ed: Feb 25, 2016. Assessed May 20, 2017).
[26] Na ional Ins i u e o Heal h and Wel a e. Ca e egis e o heal h ca e. h ps://
www. hl.fi/en/web/ hlfi-en/s a is ics/in o ma ion-on-s a is ics/ egis e -
desc ip ions/ca e- egis e - o -heal h-ca e. (Da e las upda ed: Feb 25, 2016.
Assessed May 20, 2017).
[27] R. Lie zen, P. Vi anen, M. Ki imaki, e al., S ess ul li e e en s and he onse o
as hma, Eu . Respi . J. 37 (2011) 1360–1365.
[28] J. Vah e a, M. Ki imaki, C. Hublin, e al., Liabili y o anxie y and se e e li e e en s
as p edic o s o new-onse sleep dis u bances, Sleep 30 (2007) 1537–1546.
[29] J.T. Halme, K. Seppa, H. Alho, e al., Haza dous d inking: p e alence and asso-
cia ions in he Finnish gene al popula ion, Alcohol. Clin. Exp. Res. 32 (2008)
1615–1622.
[30] B.E. Ainswo h, W.L. Haskell, M.C. Whi , e al., Compendium o physical ac i i ies:
an upda e o ac i i y codes and MET in ensi ies, Med. Sci. Spo s Exe c. 32 (2000)
S498–S504.
[31] C. Me cal e, G.D. Smi h, E. Wadswo h, e al., A con empo a y alida ion o he
Reede s ess in en o y, B . J. Heal h Psychol. 8 (2003) 83–94.
[32] A.T. Beck, R.A. S ee , M.G. Ca bin, Psychome ic p ope ies o he Beck dep ession
in en o y: wen y-fi e yea s o e alua ion, Clin. Psychol. Re . 8 (1988) 77–100.
[33] S.R. Lipsi z, K. Kim, L. Zhao, Analysis o epea ed ca ego ical da a using gene alized
es ima ing equa ions, S a . Med. 13 (1994) 1149–1163.
[34] R.J. W igh , M. Rod iguez, S. Cohen, Re iew o psychosocial s ess and as hma: an
in eg a ed biopsychosocial app oach, Tho ax 53 (1998) 1066–1074.
[35] A.H. Anis, L.D. Lynd, X. Wang, e al., Double ouble: impac o inapp op ia e use o
as hma medica ion on he use o heal h ca e esou ces, Can. Med. Assoc. J. 164
(2001) 625–631.
[36] S. Keddem, F.K. Ba g, K. Glanz, e al., Mapping he u ban as hma expe ience: using
quali a i e GIS o unde s and con ex ual ac o s affec ing as hma con ol, Soc. Sci.
Med. 140 (2015) 9–17.
[37] M. Gissle , J. Haukka, Finnish heal h and social wel a e egis e s in epidemiological
esea ch, No sk Epidemiologi 14 (2004) 113–120.
[38] J.M. Rapola, J. Vi amo, P. Ko honen, e al., Validi y o diagnoses o majo co ona y
e en s in na ional egis e s o hospi al diagnoses and dea hs in Finland, Eu . J.
Epidemiol. 13 (1997) 133–138.
[39] E.T. Nau eckas, V. Dukic, X. Bao, e al., Sho -ac ing be a-agonis p esc ip ion fills
as a ma ke o as hma mo bidi y, Ches 128 (2005) 602–608.
R. Lie zén e al. Jou nal o Psychosoma ic Resea ch 100 (2017) 46–52
52