32. King A, King D. Physical ac i i y o an aging popula ion.
Public Heal h Re 2010; 32: 401–26.
33. McMu do ME. A heal hy old age: ealis ic o u ile goal? BMJ
2000; 321: 1149–51.
34. Nooyens ACJ, Vissche TLS, Schui AJ e al.. E ec s o e i e-
men on li es yle in ela ion o changes in weigh and wais ci -
cum e ence in Du ch men: a p ospec i e s udy. Public Heal h
Nu 2005; 8: 1266–74.
35. Myck M. Physical ac i i y pa e ns o Eu opean 50+ popula-
ions. Ad Rehabil 2010; 3: 6–13.
36. Va o JJ, Ma inez-Gonzalez MA, De I ala-Es e ez J e al..
Dis ibu ion and de e minan s o seden a y li es yles in he
Eu opean Union. In J Epidemiol 2003; 32: 138–46.
37. Fo d ES, Caspe sen CJ. Seden a y beha iou and ca dio ascu-
la disease: a e iew o p ospec i e s udies. In J Epidemiol
2012; 41: 1338–53.
38. Penedo FJ, Dahn JR. Exe cise and well-being: a e iew o
men al and physical heal h benefi s associa ed wi h physical ac-
i i y. Cu Opin Psychia y 2005; 18: 189–93.
39. G an BC. ‘You’ e ne e oo old’: belie s abou physical
ac i i y and playing spo in la e li e. Ageing Soc 2001; 21:
777–98.
Recei ed 2 May 2013; accep ed in e ised o m
26 Sep embe 2013
Age and Ageing 2014; 43: 393–399
doi: 10.1093/ageing/a 192 © The Au ho 2013. Published by Ox o d Uni e si y P ess on behal o he B i ish Ge ia ics Socie y.
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/3.0/), 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]
Published elec onically 8 Decembe 2013
Job s ain in he public sec o and hospi al
in-pa ien ca e use in old age: a 28-yea
p ospec i e ollow-up
MIKAELA BIRGITTA VON BONSDORFF1,MONIKA VON BONSDORFF1,JENNI KULMALA1,TIMO TÖRMÄKANGAS1,
JORMA SEITSAMO2,PÄIVI LEINO-ARJAS2,CLAS-HÅKAN NYGÅRD3,JUHANI ILMARINEN4,TAINA RANTANEN1
1
Ge on ology Resea ch Cen e and Depa men o Heal h Sciences, Uni e si y o Jy äskylä, Finland
2
Finnish Ins i u e o Occupa ional Heal h, Helsinki, Finland
3
Ge on ology Resea ch Cen e and School o Heal h Sciences, Uni e si y o Tampe e, Finland
4
Juhani Ilma inen Consul ing, 01650 Van aa, Finland
Add ess co espondence o: M. B. on Bonsdo . Tel: +358 400 342 692; Fax: +358 14 617 422. Email: mikaela. onbonsdo @jyu. i
Abs ac
Backg ound: high job s ain inc eases he isk o heal h decline, bu li le is known abou he specific consequences and long-
e m e ec s o job s ain on old age heal h.
Objec i es: pu pose was o in es iga e whe he physical and men al job s ain in midli e was associa ed wi h hospi al ca e use
in old age.
Me hods: s udy popula ion included 5,625 Finnish public sec o employees aged 44–58 yea s who wo ked in blue- and whi e-
colla p o essions in 1981. The numbe o in-pa ien hospi al ca e days was collec ed om he Finnish Hospi al Discha ge
Regis e o he 28-yea ollow-up pe iod.
Resul s: a es o hospi al ca e days pe 1,000 pe son-yea s o men we e 7.78 (95% confidence in e al [CI] 7.71–7.84) o
low, 9.68 (95% CI 9.50–9.74) o in e media e and 12.56 (95% CI 12.47–12.66) o high physical job s ain in midli e. The co -
esponding a es o women we e 6.63 (95% CI 6.57–6.68), 7.91 (95% CI 7.87–7.95) and 10.35 (95% CI 10.25–10.42), e-
spec i ely. Ra es we e pa allel bu lowe o men al job s ain. Repo ing high physical job s ain in midli e inc eased he isk o
hospi al ca e in old age compa ed wi h hose who epo ed low job s ain, ully adjus ed incidence a e a io 1.17 (95% CI
1.00–1.38) o men and 1.42 (95% CI 1.25–1.61) o women. These associa ions we e obus in analyses confined o hospi al
ca e ha ook place a e he employees had u ned 65 yea s.
393
Job s ain and hospi al ca e use
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om
Conclusion: exposu e o high men al and, pa icula ly, high physical job s ain in midli e may se employees on a highe
heal hca e use ajec o y which pe sis s in o old age.
Keywo ds: ageing, job s ain, hospi al ca e, epidemiology, coho s udy, olde people
In oduc ion
The ising p e alence o ch onic disease and disabili y wi h
olde age inc eases he need o heal hca e se ices in he
ageing popula ion. Heal hy li ing habi s play a majo ole in
he ageing p ocess [1,2], bu o he exposu es ha he pe sons
ace du ing he li e cou se inc ease he isk o ad e se heal h
ou comes in old age. Bo h high men al and physical job s ain
con ibu e o inc eased sickness absence om wo k and ea ly
wi hd awal om he wo k o ce [3,4]. Highe men al and phys-
ical job s ain ha e been ound o inc ease he p e alence o
se e al ch onic diseases and condi ions such as co ona y hea
disease, diabe es and obesi y as well as mo ali y [5–9].
Recen findings indica e ha men al job s ain and wo k-
ela ed s ess [10–12] and physical wo k condi ions [13]ha e
long- e m e ec s on heal h and unc ioning also a e e i e-
men in old age. Al hough e idence on he long- e m e ec s
o job s ain is accumula ing, physical and men al job s ain
ha e less equen ly been in es iga ed simul aneously in a
ep esen a i e popula ion using conc e e public heal h indica-
o s such as ca e use. These findings a e o in e es while
simul aneously as he demands o he wo king li e inc ease,
he ageing wo k o ce is expec ed o ansi ion in o e i emen
a an olde age [14]. This migh po en ially exe nega i e
e ec s o he heal h and unc ioning o he indi iduals who
en e in o hi d age. In he p esen s udy, we explo e u he
he e ec s o physical and men al job s ain assessed in
midli e on he olume o hospi al in-pa ien ca e use in old
age among public sec o employees wo king in blue- and
whi e-colla p o essions.
Me hods
Pa icipan s and design
The Finnish Longi udinal S udy on Municipal Employeesby
he Finnish Ins i u e o Occupa ional Heal h included 6,257
middle-aged municipal p o essions andomly chosen om
all municipals in Finland [15,16]. A baseline in 1981, 5,625
indi iduals aged 44–58 yea s had pos al ques ionnai e da a
a ailable on physical and men al job s ain, o a esponse
a e o 89.9%. Compa ed wi h he o iginal sample, hose
wi h missing da a on job s ain in midli e we e on a e age
olde (51.5 e sus 50.3 yea s, - es P< 0.001) and mo e
o en blue-colla s (66.6% e sus 44.5%, χ
2
- es P< 0.001),
bu he e we e no di e ences in gende o heal h beha iou s
such as smoking o alcohol consump ion. The E hical
Commi ee o he Finnish Ins i u e o Occupa ional Heal h
app o ed he s udy.
Men al and physical job s ain in midli e
We used baseline da a on job s ain o a oid he ‘heal hy
wo ke e ec ’[17,18], which migh ha e esul ed in unde -
es ima ion o he associa ion be ween job s ain and hospi al
ca e use. Job s ain had mos likely been ela i ely s able
du ing he la es yea s o he wo k ca ee s, because >70% o
he male pa icipan s had held he cu en wo k posi ion o
he las 10 yea s.
Men al job s ain was examined wi h he job demand–
con ol concep desc ibed by Ka asek [19,20]. The fi e i ems
(C onbach’s alpha = 0.73) included ques ions on, e.g. p es-
su es ela ed o wo k, wo k-pace and ime schedule.
Pa icipan s we e asked whe he hese i ems we e p esen a
wo k ‘no a all = 0’,‘li le = 1’,‘somewha = 2’o ‘a lo = 3’.
The summa y sco e anged be ween 0 and 15 wi h a highe
sco e indica ing highe demand. The 10-i em scale e med
job con ol (C onbach’s alpha = 0.85) included ques ions on,
e.g. o wha ex en i was possible o influence he wo k en i -
onmen o pa icipa e in planning o he wo k. Pa icipan s
we e asked whe he he i ems we e p esen a wo k ‘no a all
=0’,‘li le = 1’,‘somewha = 2’o ‘enough = 3’.The
summa y sco e anged be ween 0 and 30 wi h a highe sco e
indica ing highe le el o job con ol. Summa y sco es on job
demand and job con ol we e calcula ed and di ided in o
hi ds indica ing low, in e media e and high le els o demand
o con ol. The men al job s ain indica o was cons uc ed
as: low job s ain (low demands wi h high o in e media e
con ol), high job s ain (high o in e media e demands and
low con ol) and in e media e job s ain (all o he combina-
ions o demands and con ol) [21].
Physical job s ain a wo k was assessed wi h h ee i ems
on ca dio espi a o y s ain [22] and h ee on musculoskele al
s ain, C onbach’s alpha = 0.87. Ca dio espi a o y s ain was
assessed wi h he ques ion: ‘How o en does physical exe -
ion in you wo k cause: (i) swea ing, (ii) b ea hlessness and
(iii) hea palpi a ion?’The answe ing al e na i es o hese
h ee i ems we e: ‘ne e o e y a ely = 0’,‘ a ely = 1’,
‘occasionally = 2’,‘qui e equen ly = 3’and ‘ equen ly = 4’.
Musculoskele al s ain was assessed wi h he ques ion ‘How
much s ain does wo k cause o you : (i) a ms, (ii) legs and
(iii) back?’The answe ing al e na i es o hese h ee i ems
we e ‘no a all/ e y li le = 0’,‘no much = 1’,‘p e y much
=3’and ‘ e y much = 4’. The summa y sco e anged
be ween 0 and 24 wi h a highe sco e indica ing highe phys-
ical s ain. The summa y sco e o hese six i ems was calcu-
la ed wi h a highe sco e indica ing highe physical s ain and
i was u he di ided in o dis ibu ion-based hi ds indica -
ing low, in e media e and high physical job s ain.
394
M. B. on Bonsdo e al.
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om
Hospi al in-pa ien ca e in old age
Hospi al in-pa ien admission and discha ge da es we e
ex ac ed om he Finnish Hospi al Discha ge Regis e
(FHDR) be ween Janua y 1, 1981 and Decembe 31, 2008
o all pa icipan s. A ecen e iew showed ha compa ed
wi h ex e nal in o ma ion on hospi al ca e, >95% o dis-
cha ges we e iden ified in he FHDR [23]. One hospi al ca e
day was defined as an o e nigh s ay o day su ge y in a
cen al, dis ic o uni e si y hospi al o heal h cen e. The
o al olume o hospi al ca e was calcula ed by adding o-
ge he all hospi al ca e days which had accumula ed du ing
he 28-yea ollow-up. Mo ali y da es we e ob ained om
he Finnish Na ional Popula ion Regis e .
Backg ound a iables
Occupa ional class was o med om he 133 di e en occu-
pa ional i les iden ified by clus e ing hem in o 13 occupa-
ions based on job analysis a he pa icipan s’wo kplaces
[24]. These we e u he collapsed in o blue-colla (e.g. main-
enance and cleaning), lowe whi e-colla ( anspo wo k
and nu sing) and uppe whi e-colla (e.g. adminis a o and
physician) p o essions. Smoking (ne e smoked, ex-smoke
o cu en smoke ), alcohol consump ion (ne e , wice a
mon h a mos o a leas once a week) and physical ac i i y
du ing p e ious yea (inac i e, mode a e ac i i y once a week
a mos o igo ous ac i i y a leas once a week) we e also
inqui ed in he ques ionnai es. Main ch onic illnesses diag-
nosed o ea ed by a physician such as ca dio ascula dis-
eases (e.g. hype ension), me abolic diso de s (e.g. diabe es)
and musculoskele al diseases (e.g. a h i is) we e inqui ed in
he ques ionnai es.
S a is ical analyses
Use o hospi al in-pa ien ca e pe 1,000 pe son-yea s was
calcula ed o he job s ain g oups indica ing he days spen
in hospi al du ing he ollow-up ha 1,000 pa icipan s in he
espec i e job s ain g oup p oduce pe yea . Follow-up ime
was calcula ed as he numbe o days be ween he beginning
o he s udy Janua y 1, 1981 and he da e o dea h o end o
ollow-up Decembe 31, 2008, whiche e happened fi s . We
excluded ca e ha ook place du ing he fi s 2 yea s o
ollow-up, in o de o a oid po en ial e e se causa ion
be ween exposu e and ou come bu because he e we e no
di e ences in hese esul s, we epo esul s o he en i e
ollow-up ime.
The associa ion be ween physical and men al job s ain in
midli e and he olume o hospi al ca e use we e es ed using
nega i e binomial eg ession models, whe e he s eng h o
an associa ion is es ima ed as an incidence a e a io (IRR)
and 95% confidence in e al (CI). IRRs a e in e p e ed as
ela i e isk es ima es and ep esen he isk o pe sons in
he p edic o a iable g oups ela i e o hose in he e e -
ence g oup. The analyses we e conduc ed o subjec s who
died o su i ed o e he ollow-up pe iod. An o se a i-
able o he leng h o ollow-up ime was used o adjus
es ima ion o a unc ion o he numbe o days he subjec s
spen unde obse a ion. Volume o hospi al ca e was a con-
inuous a iable wi h 1 day as a uni . Analyses we e adjus ed
fi s o age and leng h o ollow-up and hen o occupa ion-
al class, heal h beha iou s and main ch onic illnesses. We
analysed men and women sepa a ely (Wald es P= 0.051 o
he in e ac ion be ween gende and job s ain on hospi al
ca e use). Modelling was pe o med in he R s a is ical p o-
g amming en i onmen ( e sion 2.12.2) using he
MASS-package ( e sion 7.3–11) [25].
Fo he job s ain a iables, we impu ed missing alues
wi h a mean alue calcula ed om he o he i ems included
in he job s ain scales o hose wi h missing da a on a mos
wo i ems o fi e o job demands (2.7%), fi e i ems o 10
o job con ol (3.9%) and h ee i ems o he six o physical
job s ain (13.0%). We did comple e case analyses on he da a
bu while he esul s we e e y simila o he impu ed ones,
we show esul s om he la e analyses.
Resul s
The mean age a baseline was 50.3 yea s (SD 3.6) and 45.0%
we e men. The cha ac e is ics in midli e acco ding o gende
a e p esen ed in Table 1. O e he cou se o 28 yea s, only a
ew indi iduals (0.5%) had no in-pa ien hospi al ca e days.
Du a ion o hospi al in-pa ien ca e anged be ween 0 and
2,875 days. Median du a ion o hospi al ca e du ing he
ollow-up was 39 days (in e qua ile ange [IQR] 13, 96 days)
o men and 32 days (IQR 12, 78 days) o women. The
median numbe o hospi al ca e episodes du ing he ollow-
up was 6 (IQR 3, 12 episodes) and anged be ween 0 and
123 episodes.
O he 5,625 pa icipan s, 1,738 (1,073 men and 665
women) died du ing he ollow-up. Fo he deceased, median
du a ion o ollow-up was 19.5 yea s (IQR 13.5, 24.5 yea s)
........................................
Table 1. Cha ac e is ics o he s udy popula ion, alues a e n
(%) unless s a ed o he wise
Men n= 2,531 Women n= 3,094
Age, yea s (mean, SD) 50.4 (3.6) 50.3 (3.6)
Body mass index (mean, SD) 26.1 (3.1) 25.2 (3.5)
Occupa ional class
Uppe whi e-colla s 504 (19.9) 701 (22.7)
Lowe whi e-colla s 280 (11.1) 1,637 (52.9)
Blue-colla s 1,747 (69.0) 756 (24.4)
Ne e smoked 723 (28.7) 2,358 (76.4)
Alcohol consump ion ≥1 pe week 508 (20.2) 92 (3.0)
Vigo ous physical ac i i y ≥1 pe week 1,167 (46.7) 1,537 (50.7)
Ca dio ascula disease 594 (23.5) 630 (20.4)
Me abolic diso de 215 (8.5) 345 (11.2)
Musculoskele al disease 1,233 (39.9) 937 (37.0)
No hospi al ca e du ing he ollow-up
a
142 (0.06) 153 (0.05)
Du a ion o hospi al ca e (median, IQR)
a
39.0 (13.0, 96.0) 32.0 (12.0, 78.0)
Numbe o hospi al ca e episodes
(median, IQR)
a
6.0 (3.0, 12.0) 6.0 (3.0, 12.0)
SD, s anda d de ia ion; IQR, in e qua ile ange.
a
Calcula ed o he 28-yea ollow-up.
395
Job s ain and hospi al ca e use
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om
anging om 20 days o 28 yea s. The ones who su i ed he
ollow-up we e a baseline sligh ly younge (49.8 yea s SD 3.4
e sus 51.6 yea s SD 3.6, - es P< 0.001), mo e o en women
(78.5% e sus 57.6%, χ
2
- es P< 0.001) and whi e-colla s
(61.2% e sus 42.7%, χ
2
- es P< 0.001), epo ed mo e e-
quen ly low physical (34.3% e sus 28.8%, χ
2
- es P<0.001)
and men al (21.1% e sus 19.0%, χ
2
- es P=0.017)jobs ain
in midli e han he ones who did no su i e.
The a es o hospi al in-pa ien ca e days pe 1,000
pe son-yea s, p esen ed in Table 2, o men we e 7.78 (95%
CI 7.71, 7.84) o low, 9.68 (95% CI 9.50, 9.74) o in e -
media e and 12.56 (95% CI 12.47, 12.66) o high physical
job s ain in midli e. Fo men, hese a es we e pa allel bu
sligh ly lowe o men al job s ain. Fo women, he a es o
hospi al ca e we e 6.63 (95% CI 6.57, 6.68) o low, 7.91
(95% CI 7.87, 7.95) o in e media e and 10.35 (95% CI
10.25, 10.42) o high physical job s ain in midli e. Among
women, he e we e no di e ences in he a es o hospi al
ca e acco ding o men al job s ain in midli e.
The associa ions be ween job s ain and olume o hospi al
in-pa ien ca e in he 28-yea ollow-up a e p esen ed in
Table 3. Compa ed wi h hose public sec o employees wi h
low job s ain in midli e, men wi h high job s ain had on
a e age a 42% and women a 71% highe isk o spending an
in-pa ien day in a hospi al du ing he ollow-up, IRRs adjus ed
o age and leng h o ollow-up ime we e 1.42 (95% CI 1.24,
1.62) and 1.71 (95% CI 1.53, 1.92), espec i ely. Fo men he
associa ion be ween men al job s ain in midli e and olume o
hospi al ca e was pa allel bu lowe . Fo women, he associ-
a ion be ween men al job s ain in midli e and olume o hos-
pi al in-pa ien ca e was less clea and no s a is ically
significan . Fu he adjus men o occupa ional class, heal h
beha iou s and main ch onic illnesses assessed in midli e a e-
nua ed he IRR’s, bu hey emained s a is ically significan o
....................................................................................
Table 3. Physical and men al job s ain in midli e as p edic o s o hospi al in-pa ien ca e days in he 28-yea ollow
Men Women
IRRs (95% CI) adjus ed o IRRs (95% CI) adjus ed o
Age and leng h o
ollow-up
Age, leng h o ollow-up,
occupa ional class, heal h
beha iou and main
ch onic illnesses
Age and leng h o
ollow-up
Age, leng h o ollow-up,
occupa ional class, heal h
beha iou and main
ch onic illnesses
Sepa a e e ec s o job s ain
Physical job s ain
Low 1.00 1.00 1.00 1.00
In e media e 1.25 (1.11, 1.41) 1.11 (0.97, 1.27) 1.29 (1.16, 1.44) 1.19 (1.06, 1.33)
High 1.42 (1.24, 1.62) 1.17 (1.00, 1.38) 1.71 (1.53, 1.92) 1.42 (1.25, 1.61)
Men al job s ain
Low 1.00 1.00 1.00 1.00
In e media e 1.20 (1.05, 1.37) 1.13 (0.99, 1.30) 1.09 (0.97, 1.23) 1.09 (0.97, 1.22)
High 1.30 (1.12, 1.51) 1.14 (0.98, 1.34) 1.08 (0.94, 1.26) 0.97 (0.84, 1.12)
Simul aneous e ec s o job s ain
Physical job s ain
Low 1.00 1.00 1.00 1.00
Mode a e 1.23 (1.09, 1.39) 1.10 (0.96, 1.26) 1.30 (1.16, 1.45) 1.18 (1.05, 1.33)
High 1.38 (1.21, 1.58) 1.16 (0.99, 1.36) 1.73 (1.54, 1.95) 1.43 (1.25, 1.63)
Men al job s ain
Low 1.00 1.00 1.00 1.00
In e media e 1.17 (1.02, 1.33) 1.13 (0.99, 1.29) 1.00 (0.89, 1.13) 1.03 (0.91, 1.15)
High 1.19 (1.02, 1.39) 1.13 (0.97, 1.32) 0.93 (0.80, 1.08) 0.90 (0.77, 1.04)
....................................................................................
Table 2. Unadjus ed a es and 95% CI o hospi al in-pa ien ca e days pe 1,000 pe son-yea s acco ding o physical and men al
job s ain in midli e
Men Women
nRa e pe 1,000 pe son-yea s, (95% CI) NRa e pe 1,000 pe son-yea s, (95% CI)
Physical job s ain
Low 795 7.78 (7.71, 7.84) 1,034 6.63 (6.57, 6.68)
In e media e 1,019 9.68 (9.50, 9.74) 1,112 7.91 (7.87, 7.95)
High 717 12.56 (12.47, 12.66) 948 10.35 (10.25, 10.42)
Men al job s ain
Low 489 7.85 (7.76, 7.93) 664 8.42 (8.35, 8.49)
In e media e 1,373 10.05 (9.99, 10.11) 1,852 8.20 (8.15, 8.24)
High 669 10.89 (10.80, 10.97) 578 8.14 (8.07, 8.22)
396
M. B. on Bonsdo e al.
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om
physical job s ain. Nex , we included bo h physical and
men al job s ain simul aneously in he models o examine
whe he hey modified each o he ’s e ec on he olume o in-
pa ien hospi al ca e, see Table 3. In he model adjus ed o
age and leng h o ollow-up, he e ec s o physical and men al
job s ain in midli e on hospi al ca e in old age we e simila o
he sepa a e e ec s desc ibed abo e. Adjus ing o occupa ion-
al class, heal h beha iou s and main ch onic illnesses a enu-
a ed he associa ions pa icula ly o men.
Finally, we only included hospi al ca e ha ook place
a e he employees u ned 65 yea s (n= 5,097) o in es iga e
he long- e m e ec s o job s ain on hospi al ca e use. The
esul s we e simila , i no e en mo e p onounced, compa ed
wi h hose epo ed o all hospi al ca e du ing he ollow-up.
Discussion
We ound in his popula ion-based coho including male
and emale employees wo king in he public sec o , ha high
physical job s ain assessed in midli e p edic ed an inc eased
need o in-pa ien all-cause hospi al ca e in old age. Findings
we e pa allel bu less e iden o men al job s ain. To ou
knowledge, his is he fi s p ospec i e s udy o in es iga e
he long- e m e ec s o bo h physical and men al job s ain
on he o e all olume o hospi al ca e use in a ep esen a i e
old popula ion.
Ou s udy ex ends p e ious findings on he ela ion
be ween job s ain and mo bidi y by using hospi al in-pa ien
ca e days as a measu e o he bu den o mo bidi y o e a long
ollow-up pe iod. Ea lie findings on he e ec o highe phys-
ical job s ain on inc eased ca dio ascula mo bidi y and mo -
ali y ha e been epo ed [8,26]. A numbe o epidemiologic
s udies ha e ound ha high men al job s ain inc ease he
p e alence o ce ain ch onic diseases and condi ions such as
ca dio ascula disease, diabe es and obesi y [5,6]. Howe e ,
ew s udies ha e in es iga ed he e ec o bo h physical and
men al job s ain in he same coho and li le is s ill known
abou he e ec s o wo k s ain on old age heal h.
We ound he a e o hospi al in-pa ien ca e pe 1,000
pe son-yea s o be almos wo imes highe o hose wi h
high physical job s ain. Exposu e o high job s ain du ing he
wo k ca ee seemed o se indi iduals on a highe ca e use a-
jec o y also yea s la e , which was confi med in he p esen
analyses showing ha employees wi h high job s ain in
midli e had an inc eased isk o hospi al in-pa ien ca e also
a e u ning 65 yea s. This is in line wi h ea lie findings on
he associa ion be ween low wo k abili y in midli e and highe
p e alence o disabili y in old age [16], which is a well-known
isk ac o o inc eased need o ca e in olde age [27].
We ound ha he longe he pe son su i ed, he less
hospi al ca e days we e needed. This can be explained by he
inc ease in hospi al ca e need in he las yea s o li e due o
dec easing heal h [28]. Con a y o he common iew o
dea h being a compe ing isk, he compe ing isk in ou
s udy was he end o he ollow-up. This mean ha mos o
he pa icipan s who su i ed un il he end o he ollow-up
did no each he las yea s o li e whe e need o hospi al
ca e is highes . Fu he s udies in unselec ed deceden popu-
la ions a e wa an ed o examine job s ain as a p edic o o
hospi al ca e use in he las yea s o li e.
The s udy s eng hs include a long p ospec i e ollow-up in
a la ge popula ion-based da ase consis ing o a wide a ie y o
p o essions. We had comple e egis e -based da a a ailable
un il old age o all pa icipan s o he en i e s udy pe iod and
did no lose anyone o ollow-up. The e a e some limi a ions
ha need o be add essed. Fi s , we impu ed missing alues in
he job s ain i ems a midli e which migh ha e in oduced
some bias o he esul s. Howe e , he esul s we e simila
when using non-impu ed da a. Second, he pa icipan s wi h
da a on job s ain we e younge and mo e o en whi e-colla s
han hose who did no ha e da a a ailable on job s ain.
Fu he mo e, ou s udy included occupa ionally ac i e public
sec o employees and should be conside ed when gene alising
he p esen esul s on a popula ion le el.
In ou s udy, high men al and pa icula ly physical job s ain
in midli e inc eased he olume o hospi al in-pa ien ca e use
in old age. These findings a e o clinical impo ance while he
demands o he wo king li e inc ease among he ageing wo k-
o ce, which is expec ed o con inue o wo k in olde age in
Wes e n coun ies. A ecen me a-analysis [29] showed ha a
heal hy li es yle educed he isk o co ona y a e y disease sub-
s an ially among hose wi h high job s ain. This indica es ha
along wi h measu es ha imp o e wo k condi ions and de-
c ease job s ain among he middle-aged employees, p omo ing
a heal hy li es yle is likely o educe he long- e m nega i e
e ec s o job s ain. Fo example, wo k place physical ac i i y
p og ams ha e yielded posi i e esul s in e ms o inc eased
physical ac i i y and dec eased musculoskele al diso de s [30].
Howe e , mo e esea ch on he e ec i eness o p og ams ha
help p ese e he heal h and unc ioning o ageing employees
is wa an ed in he e o o p e en inc ease in he olume o
heal hca e se ices needed in old age.
Key poin s
•P e ious s udies show ha bo h high men al and physical
job s ain con ibu e o ea ly wi hd awal om he wo k-
o ce and inc ease he p e alence o se e al ch onic dis-
eases, bu li le is s ill known abou he specific
consequences (e.g. hospi al ca e) and long- e m e ec s o
job s ain on old age heal h.
•Ou s udy ound ha high men al and pa icula ly physical
job s ain in midli e inc eased he olume o hospi al in-
pa ien ca e use in old age in he long- e m ollow-up.
•The associa ions we e obus also in analyses confined o
hospi al ca e ha ook place a e he employees had u ned
65 yea s.
•Exposu e o high men al and, pa icula ly, high physical job
s ain in midli e may se employees on a highe heal hca e
use ajec o y which pe sis s in o old age.
397
Job s ain and hospi al ca e use
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om

Funding
FLAME was suppo ed by he Finnish Ins i u e o
Occupa ional Heal h and he Local Go e nmen s Pension
Ins i u ion. MB B was suppo ed by he Academy o
Finland g an numbe s 132597 and 257239. ME B was sup-
po ed by he Academy o Finland g an numbe 250681. JK
was suppo ed by he Academy o Finland g an 250385. TT
was suppo ed by he Academy o Finland g an numbe
132597.
Con lic s o in e es
None decla ed.
Re e ences
1. Sabia S, Singh-Manoux A, Hagge -Johnson G, Cambois E,
B unne EJ, Ki imäki M. Influence o indi idual and com-
bined heal hy beha iou s on success ul aging. CMAJ 2012;
184: 1985–92.
2. Chak a a y EF, Hube HB, K ishnan E, B uce BB, Lingala VB,
F ies JF. Li es yle isk ac o s p edic disabili y and dea h in
heal hy aging adul s. Am J Med 2012; 125: 190–7.
3. Laine S, Gimeno D, Vi anen M e al. Job s ain as a p edic o
o disabili y pension: he Finnish public sec o s udy.
J Epidemiol Communi y Heal h 2009; 63: 24–30.
4. Duij s SF, Kan I, Swaen GM, an den B and PA, Zeege s
MP. A me a-analysis o obse a ional s udies iden ifies p edic-
o s o sickness absence. J Clin Epidemiol 2007; 60: 1105–15.
5. Ki imäki M, Nybe g ST, Ba y GD e al. Job s ain as a isk
ac o o co ona y hea disease: a collabo a i e me a-analysis
o indi idual pa icipan da a. Lance 2012; 380: 1491–7.
6. He aclides A, Chandola T, Wi e DR, B unne EJ.
Psychosocial s ess a wo k doubles he isk o ype 2 diabe es
in middle-aged women: e idence om he Whi ehall II s udy.
Diabe es Ca e 2009; 32: 2230–5.
7. Nybe g ST, Heikkila K, F ansson EI e al. Job s ain in ela ion
o body mass index: Pooled analysis o 160 000 adul s om 13
coho s udies. J In e n Med 2012; 272: 65–73.
8. Bou gka d E, Wild P, Massin N e al. Associa ion o physical
job demands, smoking and alcohol abuse wi h subsequen p e-
ma u e mo ali y: a 9-yea ollow-up popula ion-based s udy. J
Occup Heal h 2008; 50: 31–40.
9. Hol e mann A, Mo ensen OS, Sogaa d K, Gyn elbe g F,
Suadicani P. Risk ac o s o ischaemic hea disease mo ali y
among men wi h di e en occupa ional physical demands. A
30-yea p ospec i e coho s udy. BMJ Open 2012; 2:
e000279.
10. on Bonsdo MB, Sei samo J, on Bonsdo ME, Ilma inen J,
Nygå d CH, Ran anen T. Job s ain among blue-colla and
whi e-colla employees as a de e minan o o al mo ali y: a
28-yea popula ion-based ollow-up. BMJ Open 2012; 2:
e000860.
11. Kulmala J, on Bonsdo MB, S enholm S e al. Pe cei ed
s ess symp oms in midli e p edic disabili y in old age: a
28-yea p ospec i e coho s udy. J Ge on ol A Biol Sci Med
Sci 2013 doi: 10.1093/ge ona/gls339.
12. Wah endo M, Sembajwe G, Zins M, Be kman L, Goldbe g M,
Sieg is J. Long- e m e ec s o psychosocial wo k s ess in midli e
on heal h unc ioning a e labo ma ke exi - esul s om he
GAZEL s udy. J Ge on ol B Psychol Sci Soc Sci 2012;
67: 471–80.
13. Pla s LG, Ne u eli G, Webb E e al. Physical occupa ional
exposu es du ing wo king li e and quali y o li e a e labou
ma ke exi : Resul s om he GAZEL s udy. Aging Men
Heal h 2013 doi: 10.1080/13607863.2013.781120.
14. Uni ed Na ions, Depa men o Economic and Social A ai s,
Popula ion Di ision. Wo ld Popula ion P ospec s: The 2010
Re ision, Highligh s and Ad ance Tables. Wo king Pape No.
ESA/P/WP.220. Uni ed Na ions; 2011.
15. Ilma inen J, Tuomi K, Eskelinen L, Nygå d CH, Huuh anen P,
Klocka s M. Backg ound and objec i es o he Finnish e-
sea ch p ojec on aging wo ke s in municipal occupa ions.
Scand J Wo k En i on Heal h 1991; 17(Suppl. 1): 7–11.
16. on Bonsdo MB, Sei samo J, Ilma inen J, Nygå d CH, on
Bonsdo ME, Ran anen T. Wo k abili y in midli e as a p e-
dic o o mo ali y and disabili y in la e li e: A 28-yea p o-
spec i e ollow-up s udy. CMAJ 2011; 183: E235–42.
17. Ki imäki M, Theo ell T, Wes e lund H, Vah e a J, Al edsson L.
Job s ain and ischaemic disease: does he inclusion o olde
employees in he coho dilu e he associa ion? The WOLF
S ockholm s udy. J Epidemiol Communi y Heal h 2008; 62:
372–4.
18. Theo ell T, Tsu sumi A, Hallquis J e al. Decision la i ude, job
s ain, and myoca dial in a c ion: a s udy o wo king men in
S ockholm. The SHEEP s udy g oup. S ockholm hea epi-
demiology p og am. Am J Public Heal h 1998; 88: 382–8.
19. Ka asek RA. Job demands, job decision la i ude, and men al
s ain: implica ions o job edesign. Adminis Sci Qua 1979;
24: 285–308.
20. Ka asek RA, Theo ell T. Heal hy Wo k. S ess, P oduc i i y,
and he Recons uc ion o Wo king Li e. New Yo k: Basic
Books, 1990.
21. Ki imäki M, Leino-A jas P, Luukkonen R, Riihimaki H,
Vah e a J, Ki jonen J. Wo k s ess and isk o ca dio ascula
mo ali y: p ospec i e coho s udy o indus ial employees.
BMJ 2002; 325: 857.
22. Eskelinen L, Toikkanen J, Tuomi K, Mauno I, Nygå d CH,
Ilma inen J. Symp oms o men al and physical s ess in di e -
en ca ego ies o municipal wo k. Scand J Wo k En i on
Heal h 1991; 17(Suppl. 1): 82–6.
23. Sund R. Quali y o he Finnish hospi al discha ge egis e : a
sys ema ic e iew. Scand J Public Heal h 2012; 40: 505–15.
24. Ilma inen J, Suu näkki T, Nygå d CH, Landau K.
Classifica ion o municipal occupa ions. Scand J Wo k En i on
Heal h 1991; 17(Suppl. 1): 12–29.
25. R De elopmen Co e Team. R: A language and en i onmen
o s a is ical compu ing. Vienna, Aus ia: R Founda ion o
S a is ical Compu ing, 2011.
26. Vi kkunen H, Hä mä M, Kauppinen T, Tenkanen L. Shi
wo k, occupa ional noise and physical wo kload wi h ensuing
de elopmen o blood p essu e and hei join e ec on he
isk o co ona y hea disease. Scand J Wo k En i on Heal h
2007; 33: 425–34.
27. F ied TR, B adley EH, Williams CS, Tine i ME. Func ional
disabili y and heal h ca e expendi u es o olde pe sons. A ch
In e n Med 2001; 161: 2602–7.
28. on Bonsdo MB, Ran anen T, Leinonen R, Kujala UM,
Tö mäkangas T, Män y M e al. Physical ac i i y his o y and
end-o -li e hospi al and long- e m ca e. J Ge on ol A Biol Sci
Med Sci 2009; 64: 778–84.
398
M. B. on Bonsdo e al.
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om
29. Ki imäki M, Nybe g ST, F ansson EI e al. Associa ions o job
s ain and li es yle isk ac o s wi h isk o co ona y a e y
disease: a me a-analysis o indi idual pa icipan da a. CMAJ
2013; 185: 763–9.
30. P ope KI, Koning M, an de Beek AJ, Hildeb and VH,
Bossche RJ, an Mechelen W. The e ec i eness o wo ksi e
physical ac i i y p og ams on physical ac i i y, physical fi ness,
and heal h. Clin J Spo Med 2003; 13: 106–17.
Recei ed 2 May 2013; accep ed in e ised o m
26 Sep embe 2013
Age and Ageing 2014; 43: 399–405
doi: 10.1093/ageing/a 183 © The Au ho 2013. Published by Ox o d Uni e si y P ess on behal o he B i ish Ge ia ics Socie y.
All igh s ese ed. Fo Pe missions, please email: [email p o ec ed]
Published elec onically 25 No embe 2013
Do heal h p o ide –pa ien ela ionships ma e ?
Explo ing den is -pa ien ela ionships and o al
heal h- ela ed quali y o li e in olde people
VANESSA ELAINE MUIRHEAD1,WAGNER MARCENES1,DESMOND WRIGHT2
1
Cen e o Clinical and Diagnos ic O al Sciences, Ins i u e o Den is y, Ba s and he London School o Medicine and Den is y,
Queen Ma y Uni e si y o London, UK
2
Public Heal h England, London Region, London, UK
Add ess co espondence o: V. E. Mui head. Tel: 0207 882 8637; Fax: 0207 882 5842. Email: .mui [email protected]
Abs ac
Backg ound: pa ien expe ience is now a key pa ame e in heal h ca e. Ye , e y li le is known abou he possible impac o
den is –pa ien ela ionships on pa ien -cen ed ou comes including olde peoples’o al heal h- ela ed quali y o li e
(OHRQoL).
Objec i e: his s udy assessed he ela ionship be ween OHRQoL and den is –pa ien ela ionships ela ed o pe cei ed
unme den al needs; sha ed decision-making; ime spen discussing o al heal h p oblems; espec and confidence and us .
Pa icipan s: olde people aged 65 yea s and o e li ing in Eas London, UK in 2011.
Me hods: a c oss-sec ional s udy using s a ified andom sampling ec ui ed a ep esen a i e sample o olde people (n= 772).
Pa icipan s comple ed an o al examina ion and a s uc u ed ques ionnai e including he O al Heal h Impac P ofile-14
(OHIP-14) measu ing OHRQoL and fi e den is –pa ien ela ionship ques ions aken om he UK 2009 Adul Den al Heal h
Su ey. Mul i a ia e Poisson eg essions modelled he associa ion be ween OHRQoL and den is –pa ien ac o s adjus ing o
socio-demog aphic ac o s, clinical o al indica o s, and den al a endance.
Resul s: ha ing a pe cei ed unme need o den al ea men (PRR = 1.84; 95% CI: 1.32, 2.56) and exp essing a lack o us
and confidence in one’s den is (PRR = 1.74; 95% CI: 1.01, 2.98) we e significan p edic o s o poo OHRQoL among olde
people.
Conclusions: hese findings sugges ha olde people wi h unme den al needs and hose who exp essed a lack o us and
confidence in hei den is we e mo e likely o expe ience poo OHRQoL ein o cing he impo ance o he den al pa ien
expe ience in heal hy ageing and well-being.
Keywo ds: o al heal h, den is s, quali y o li e, pa ien –p o ide ela ionships, olde people
In oduc ion
D ama ic imp o emen s in o al heal h o e he pas 50 yea s
in mos indus ialised coun ies now mean ha mo e people
e ain hei na u al ee h in o olde age [1]. UK Adul O al
Heal h Su eys ha e shown a decline in he pe cen age o
olde people wi h no na u al ee h om 78% in 1978 o only
24% in 2009 [2]. The p e e ence o keeping ee h is socially
399
Do heal h p o ide –pa ien ela ionships ma e ?
a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Augus 9, 2016h p://ageing.ox o djou nals.o g/Downloaded om