Occupational variation in incidence of bladder cancer: a comparison of population-representative cohorts from Nordic countries and Canad
Full text
1
HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
Occupa ional a ia ion in incidence o
bladde cance : a compa ison o
popula ion- ep esen a i e coho s om
No dic coun ies and Canada
Kisho Hadkhale,1 Jill MacLeod,2 Paul A Deme s,2,3,4 Jan I a Ma insen,5
Elisabe e Weide pass,5,6,7,8 K is ina Kjae heim,5 Elsebe h Lynge,9 Pä Spa en,8
Lau ey T ygg ado i ,10,11 M Anne Ha is,2,3,12 Michael Tjepkema,13 Paul A Pe e s,14
Ee o Pukkala1,15
To ci e: HadkhaleK, MacLeodJ,
Deme sPA, e al. Occupa ional
a ia ion in incidence o
bladde cance : a compa ison
o popula ion- ep esen a i e
coho s om No dic coun ies
and Canada. BMJ Open
2017;7:e016538. doi:10.1136/
bmjopen-2017-016538
►P epublica ion his o y o
his pape is a ailable online.
To iew hese iles please isi
he jou nal online (h p:// dx. doi.
o g/ 10. 1136/ bmjopen- 2017-
016538).
Recei ed 21 Feb ua y 2017
Re ised 2 June 2017
Accep ed 12 June 2017
Fo numbe ed a ilia ions see
end o a icle.
Co espondence o
D Kisho Hadkhale;
kisho . hadkhale@ s a . u a. i
Resea ch
Abs AC
Objec i es The objec i e o his s udy was o compa e
occupa ional a ia ion o he isk o bladde cance in he
No dic coun ies and Canada.
Me hods In he No dic Occupa ional Cance s udy
(NOCCA), 73 653 bladde cance cases we e obse ed
du ing ollow-up o 141.6 million pe son-yea s. In
he Canadian Census Heal h and En i onmen Coho
(CanCHEC), 8170 cases we e obse ed du ing he ollow-
up o 36.7 million pe son-yea s. S anda dised incidence
a ios wi h 95% CI we e es ima ed o 53 occupa ions in
he NOCCA coho and HR wi h 95% CIs we e es ima ed
o 42 occupa ions in he CanCHEC.
esul s Ele a ed isks o bladde cance we e obse ed
among hai d esse s, p in e s, sales wo ke s, plumbe s,
pain e s, mine s and laund y wo ke s. Teache s and
ag icul u al wo ke s had educed isk o bladde cance
in bo h coho s. Chimney-sweeps, obacco wo ke s and
wai e s had abou 1.5- old isk in he No dic coun ies; no
isk es ima es o hese ca ego ies we e gi en om he
CanCHEC coho .
Conclusion We obse ed di e en occupa ional pa e ns
in isk o bladde cance in No dic coun ies and Canada.
The only occupa ion wi h simila ly inc eased isk was
obse ed among sales wo ke s. Di e ences in smoking
ac oss occupa ional g oups may explain some, bu no all,
o his a ia ion.
In OduC IOn
Bladde cance is he nin h mos common
cance in he wo ld and occupa ion has been
iden i ied as he mos impo an isk ac o
second o smoking.1 I has been es ima ed
ha occupa ional exposu e could accoun o
as much as 20% o all bladde cance s diag-
nosed in indus ialised coun ies.1 2 In he
No dic coun ies, bladde cance is he i h
mos common cance in men and 15 h mos
common cance in women.3 Age-s anda dised
incidence a es o bladde cance inc eased
un il 1990, wi h he highes incidence in
Denma k and lowes in Finland.4 In Canada,
bladde cance is he ou h mos common
cance in men and 12 h mos common
cance in women. The e has been li le o no
change in bladde cance incidence among
bo h male and emale Canadians in he las
decade.5
Associa ions wi h bladde cance ha e
been obse ed o >40 occupa ions. While
some indings ha e been consis en ,
o he s ha e been inconsis en o limi ed.6 7
In-dep h s udies wi h high s a is ical powe
a e equi ed o es ablish clea associa ions
o occupa ional isks o bladde cance .8
Fu he mo e, occupa ional bladde
cance isk ac o s may ha e changed o e
ime, sugges ing a need o mo e cu en
e idence.9 10 Ou side o he No dic coun-
ies, only ew coun ies ha e di ec ly linked
in o ma ion on cance and census da a. The
a ailabili y o hese wo coho s conduc ed
in di e en a eas o he wo ld p o ided
s eng hs and limi a ions o his s udy
►Only ew coun ies ha e di ec ly linked in o ma ion
on cance and census da a and his s udy p o ides
in-dep h analysis wi h high s a is ical powe o
popula ion-le el compa ison using linkage da a in
No dic coun ies and Canada.
►The a ailabili y o hese wo la ge coho s and long
du a ion o ollow-up conduc ed in di e en a eas
o he wo ld p o ided an oppo uni y o examine
occupa ional a ia ion o bladde cance isk in
di e en geog aphic egions.
►The lack o da a on smoking in o ma ion and
wo kplace exposu e o chemical sol en s in his
s udy would be e explain he isk associa ed wi h
bladde cance .
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
2HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
an oppo uni y o examine occupa ional a ia ion o
bladde cance isk in No dic and Canadian popula-
ion-based coho s.
MA e IAls And Me hOds
The p esen s udy is based on he coho de i ed om he
No dic Occupa ional Cance s udy (NOCCA) ollowed
up om 1961 o 2005 and he Canadian Census Heal h
and En i onmen Coho (CanCHEC) ollowed up om
1991 o 2005 de i ed om he Canadian Census Mo ali y
and Cance Follow-up S udy ( igu e 1).
The NOCCA coho comp ises indi iduals in Denma k,
Finland, Iceland, No way and Sweden who pa icipa ed
in one o mo e popula ion censuses in 1960, 1970, 1980–
1981 o 1990, who we e be ween he ages o 30 and 64
yea s and li ing in he coun y on 1 Janua y a e he
census. The p esen s udy used he da a om he coho
ollowed up only om he pe iod 1991 o 2005 o inc ease
compa abili y wi h he CanCHEC s udy. The census
eco ds include ques ionnai e-based in o ma ion on
economic ac i i y, occupa ion and indus y, which we e
cen ally coded and compu e ised in he na ional s a is-
ics o ices. Occupa ions we e coded in o >300 ca ego ies
acco ding o he na ional adap a ions o he In e na ional
S anda d Classi ica ion o Occupa ions (ISCO) om 1958
in Finland, No way and Sweden. In Denma k, special
na ional nomencla u e was used wi h a dis inc ion om
sel -employed pe sons, amily wo ke s, sala ied employees,
skilled wo ke s and unskilled wo ke s. In Iceland, i was
coded acco ding o he ISCO-68 classi ica ion and la e
con e ed in o ISCO-58 wi h ins uc ions om In e na-
ional Labou O ganisa ion. Fo he NOCCA s udy, he
o iginal na ional codes we e con e ed acco ding o he
No dic occupa ional classi ica ion (NYK) in o 53 ca ego-
ies wi h 1 addi ional ca ego y o economically inac i e
pe sons. NYK is he No dic adap a ion o ISCO om 1958
(de ails o he occupa ional ca ego ies in NOCCA s udy
a e a ailable a : h p:// as a. cance . i/ NOCCA/ ca ego-
ies. h ml).
Da a on inciden cance cases we e ob ained om he
na ional cance egis ies in each o he No dic coun ies.
These egis ies cap u e in o ma ion om clinical and
pa hological depa men s, gene al p ac i ione s, p i a e
clinics and dea h egis e s ha a e ai ly simila in all
No dic coun ies.11 All he pa ien s had bladde cance
a he age o diagnosis. De ails o he NOCCA s udy ha e
been desc ibed elsewhe e.12 NOCCA esul s a e p esen ed
in s anda dised incidence a ios (SIRs) o es ima e isks o
bladde cance ac oss occupa ional g oups wi h a es o
he en i e na ional s udy popula ions used as he e e -
ence a es. The SIR was calcula ed as he a io o obse ed
o expec ed cases. Exac 95% CIs we e de ined based on
a Poisson dis ibu ion.
The CanCHEC coho was de i ed om esponden s
o he 1991 Canadian Census who we e included in he
Canadian Census Mo ali y and Follow-Up S udy.13 The
p esen s udy used da a om he linkage o he 1991 Cana-
dian Census 2B (long o m) wi h he Canadian Mo ali y
Da abase (1991–2011), Canadian Cance Regis y (1992–
2010) and His o ical Tax Summa y Files (1984–2011).
The na ionally ep esen a i e coho included 2 735 152
indi iduals o he Canadian non-ins i u ional esiden
popula ion aged om 25 o 74 yea s on he census day
(4 June 1991), who we e esiden s o Canada and among
he 20% o Canadian household selec ed o comple e
he long- o m census ques ionnai e. Occupa ion, coded
acco ding o he 1991 S anda d Occupa ional Classi ica-
ion (SOC-91), and socioeconomic cha ac e is ics we e
ob ained om he census. The majo i y o he occupa-
ion g oupings we e compa able be ween CanCHEC and
NOCCA. Howe e , some g oups such as chimney-sweeps
appea ed only in NOCCA while o he s, such as wai e s,
we e con ained wi hin a b oade occupa ional g oup
o ood and be e age wo ke s and obacco wo ke s in a
b oade g oup o ood p ocessing wo ke s. The CanCHEC
was ollowed up o cance mo bidi y h ough linkage o
he Canadian Cance Regis y da abase h ough de e -
minis ic and p obabilis ic me hods.
Pe son- ime a isk was coun ed om he coho en y
on 4 June 1991 o da e o disease diagnosis, dea h and loss
o ollow-up o end o ollow-up on 31 Decembe 2010,
whiche e occu ed i s . Fo he pu pose o analyses,
he i s inciden p ima y bladde cance was consid-
e ed and g oups we e e-coded based on he ou -digi
SOC codes. De ails o he CanCHEC s udy ha e been
Figu e 1 Flowcha o he me hodological p ocedu e in
NOCCA coho and CanCHECs. CanCHEC, he Canadian
Census Heal h and En i onmen Coho ; NOCCA, he No dic
Occupa ional Cance s udy.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
3
HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
desc ibed elsewhe e.13 Cox p opo ional HRs and 95%
CIs we e calcula ed using Cox p opo ional haza ds
disease- ee su i al analysis o es ima e isks o bladde
cance associa ed wi h employmen acco ding o occupa-
ion a baseline. In he absence o in o ma ion ega ding
po en ial con ounde s such as li es yle ac o s, including
smoking, physical ac i i y and die , models we e adjus ed
o age g oup, egion and le el o educa ion. In acco -
dance wi h S a is ics Canada disclosu e guidelines, no
coun s <5 o co esponding model ou pu s a e epo ed,
and all pe son-yea s and coun s a e andomly ounded
o base 5. Resul s may also be supp essed whe e coun s
<5 would be iden i ied due o addi i i y ac oss subg oups.
STROBE guidelines was ollowed o epo each sec ions,
whe e applicable.
Demog aphic in o ma ion and isk o bladde cance
by occupa ional ca ego y a e p esen ed by sex o he
NOCCA coho and he CanCHEC. Due o he absence
o in o ma ion on indi idual smoking beha iou , we also
examined he isk o lung cance and i s ela ionship wi h
he isk o bladde cance in men ac oss occupa ional
g oups ( igu e 2).
e hICAl COnsIde A IOns
E hical commi ees and da a inspec ion boa ds om
each No dic coun y app o ed he NOCCA s udy. The
CanCHEC s udy was app o ed by he Uni e si y o
To on o Heal h Sciences Resea ch E hics Boa d.
esul s
In he NOCCA coho , 73 653 cases o bladde cance
we e diagnosed among 14 902 573 (50.0% men) indi id-
uals o e he 1991–2005 ollow-up pe iod (141.6 million
pe son-yea s) ( able 1). In he CanCHEC, 8170 cases o
bladde cance we e diagnosed among 2 051 315 (54.0%
men) indi iduals du ing he 1991–2010 ollow-up pe iod
(36.7 million pe son-yea s).
The g ea es s a is ically signi ican excess isks o
bladde cance o male wo ke s in he NOCCA coho
we e obse ed among obacco wo ke s (1.67), chim-
ney-sweeps (1.47) and wai e s (1.41), occupa ions ha
we e no examined in CanCHEC ( able 2). Lowe bu
s a is ically signi ican ele a ed isks in NOCCA we e also
obse ed among sales wo ke s, p in e s, me al wo ke s
and pain e s. Among hem, only sales wo ke s showed an
ele a ed HR in CanCHEC. In u n, he e we e inc eased
isks in he CanCHEC among hai d esse s, welde s and
adminis a o s and manage s, while he SIRs o hese
occupa ions we e close o 1.0 excep o hai d esse s in
he NOCCA s udy ( able 2). The co ela ion coe icien
( ) o lung and bladde cance in men in NOCCA and
CanCHEC s udy was obse ed as 0.75 and 0.81, espec-
i ely ( igu e 2).
The g ea es s a is ically signi ican excess isks o
emale wo ke s in he NOCCA coho we e obse ed
among obacco wo ke s (1.65), occupa ion ha was no
examined sepa a ely in CanCHEC ( able 3). S a is ically
signi ican excess isk in CanCHEC was obse ed only
among plumbe s. S a is ically signi ican ele a ed isks
in NOCCA we e also obse ed among p in e s, chemical
wo ke s, hai d esse s, wai e s, sales wo ke s and admin-
is a i e manage s. While in CanCHEC, an insigni ican
inc eased isk was obse ed among p in e s, hai d esse
and sales wo ke s. Wai e s we e no examined sepa-
a ely in he CanCHEC s udy ( able 3). Reduced isks
we e consis en ly obse ed o eache s and ag icul u e
wo ke s in bo h coho s.
dIsCussIOn
This compa ison o esul s om wo la ge coho s udies
e ealed occupa ional a ia ion in bladde cance isks.
Figu e 2 Co ela ion o bladde and lung cance in men in NOCCA coho and CanCHEC. AM, adminis a o s and manage s;
AW, ag icul u e wo ke s; CanCHEC, Canadian Census Heal h and En i onmen Coho ; CH, chemical wo ke s; CS, chimney-
sweeps; CW, cons uc ion wo ke s; DV, d i e s; EW, elec ical wo ke s; FB, ood and be e age wo ke s; FF, i e igh e s; FP,
ood-p ocessing wo ke s; FW, o es y wo ke s; HD, hai d esse s; LW, laund y wo ke s; MN, mine s; MW, me al wo ke s;
NOCCA; No dic Occupa ional Cance s udy; PA, pain e s; PI, p in e s; PL, plumbe s; RP, ubbe and plas ics wo ke s; SIR,
s anda dised incidence a io; SW, sales wo ke s; TH, eache s; TW, obacco wo ke s; WA, wai e s; WE, welde s.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
4HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
Gene ally, he excesses in a gi en occupa ion seen in one
coho we e no seen in he o he . Male hai d esse s in
CanCHEC and emale hai d esse s in NOCCA showed
a ma ked excess isk o bladde cance while he e was
almos no indica ion o an excess isk among emale hai -
d esse s in CanCHEC and non-signi ican inc eased isk
among male hai d esse s in NOCCA. The inding o an
excess isk is consis en wi h wo di e en me a-analyses:
(1) Ha ling and colleagues14 epos ed a summa y isk
a io (SRR) o 1.70 (95% CI 1.10 o 2.88) and (2) Takk-
ouche and colleagues,15 a isk a io (RR) o 1.30 (95%
CI 1.20 o 1.42). Hai d esse s a e exposed o chemical
compounds o a oma ic amines such as benzidine, olu-
idine and a oma ic ni ous compounds, and exposu e
o hese chemical compounds has been linked o an
inc eased isk o bladde cance .16 In pa icula , long- e m
exposu e (≥10 yea s) as a hai d esse has been associa ed
wi h g ea e isk o bladde cance .14 17
Exposu es o ca cinogenic chemicals among hai -
d esse s appea o ha e been diminishing in ecen
decades.18 Following he ban on use o a oma ic amines
in he Eu opean Union in 1980s, he ea e a simila ban
was also imposed in Canada. Some mo e ecen s udies
ha examined exposu es a e hese bans ha e no
epo ed excess isks o bladde cance , bu some s udies
ha e epo ed ha banned subs ances a e s ill in use,
leading o he po en ial o ongoing exposu e.19 The pe i-
od-speci ic s a i ied analysis o he NOCCA s udy om
1991 o 2005 also shows a simila dec ease in isk among
hai d esse s o e ime in he No dic coun ies and in
Canada ( esul s no shown). As ou coho s ep esen he
ollow-up om 1991, he highes isks obse ed in 1990s
may ha e been due o he use o a oma ic amines in dye
p oduc s, which a e no used any mo e.19
We obse ed some e idence o an associa ion be ween
bladde cance and employmen as a p in e in his
compa a i e s udy, wi h he g ea es excess isk obse ed
among No dic women. A me a-analysis o case-con ol
s udies om six Eu opean coun ies obse ed ha occu-
pa ion ela ed o p in ing had an up o 80% g ea e isk
o bladde cance han he gene al popula ion (OR 1.81,
95% CI 1.03 o 3.17).20 A Spanish s udy also obse ed a
signi ican ly ele a ed isk o bladde cance among men
in p in ing indus y (OR 5.4, 95% CI 1.62 o 17.72).21
P in e s a e mainly exposed o p in ing inks consis ing o
pigmen s and a sol en .22 Exposu es o chemical agen s
Table 1 Demog aphic and o he a iables o he s udy popula ion in No dic coun ies (NOCCA) and Canada (CanCHEC)
coho s
Va iable
NOCCA CanCHEC
Ca ego y Numbe (%) Ca ego y Numbe (%)
Popula ion by coun y/
p o ince
NOCCA (1961–2005)
CanCHEC (1991–2010)
Denma k
Finland
Iceland
No way
Sweden
2 013 346 (13.5%)
3 404 800 (22.8%)
120 995 (0.8%)
2 562 674 (17.2%)
6 800 758 (45.6%)
A lan ic
Quebec
On a io
Mani oba
Saska chewan
Albe a
B i ish Columbia
Yukon, NW Te i o y and
Nuna u
162 135 (7.9%)
502 495 (24.5%)
756 295 (36.8%)
87 565 (4.3%)
77 855 (3.8%)
200 135 (9.7%)
243 735 (11.8%)
21 100 (1.2 %)
Popula ion by sex Male 7 447 726 (49.97%) 1 108 410 (54.03%)
Female 7 454 847 (50.03%) 942 905 (45.96%)
Bladde cance cases
by coun y/p o ince
NOCCA (1991–2005)
CanCHEC (1991–2010)
Denma k
Finland
Iceland
No way
Sweden
17 525 (23.79%)
11 109 (15.08%)
573 (0.78%)
13 798 (18.73%)
30 648 (41.61%)
A lan ic
Quebec
On a io
Mani oba
Saska chewan
Albe a
B i ish Columbia
Yukon, NW e i o y and
Nuna u
686 (8.4%)
2356 (28.8%)
2436 (29.8%)
350 (4.3%)
383 (4.7%)
839 (10.3%)
1082 (13.2%)
42 (0.5%)
Bladde cance cases
by age g oup (yea s)
30–34
35–44
45–54
55–64
65–74
75+
15 (0.02%)
506 (0.69%)
3546 (4.82%)
12 241 (16.62%)
25 258 (34.29%)
32 087 (43.56%)
25–34
35–44
45–54
55–64
65–74
396 (4.8%)
1365 (16.7%)
2558 (31.3%)
2972 (36.4)
879 (10.7%)
Bladde cance cases
by sex
Male 54 737 (74.32%) 6630 (81.1%)
Female 18 916 (25.68%) 1540 (18.8%)
CanCHEC, Canadian Census Heal h and En i onmen Coho ; NOCCA, No dic Occupa ional Cance s udy.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
5
HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
in he p in ing indus y ha e been associa ed wi h an
inc eased isk o bladde cance .21
Small ye consis en excess isks o bladde cance we e
obse ed o sales wo ke s o bo h sexes in he Canadian
and No dic coho s. This associa ion has been obse ed
p e iously o men in a Swedish coho s udy.10 In he
Swedish s udy, physical inac i i y and seden a y wo k
we e sugges ed as he p obable cause o inc eased isk o
bladde cance . The indings in his Swedish s udy and he
p esen s udy a e no independen due o ma ked o e lap
in he da a be ween hese s udies. The me a-analysis by
’ Manne je and Pea ce23 epo ed a smoking-adjus ed
excess isk o bladde cance among emale sales wo ke s.
The s udy obse ed posi i e causal associa ion wi h
du a ion o employmen among sales wo ke s. Lowe
equency o u ina ion and educed luid in ake we e
plausible explana ions o he isk o bladde cance in
his g oup.23
Ele a ed isk o bladde cance was obse ed among
emale, bu no male, d i e s in ou compa a i e s udy. A
me a-analysis o he occupa ional isk o bladde cance
s udy obse ed a signi ican ly inc eased isk o bladde
cance among bus d i e s wi h SRR o 1.29 (95% CI 1.08
o 1.53).8 Simila ly, a signi ican posi i e dose– esponse
ela ionship be ween he du a ion o employmen as a
uck d i e and he isk o bladde cance was obse ed in
a s udy by Sil e man and colleagues.7 This excess isk may
be a ibu able o exposu e o se e al ai pollu an s such
as polycyclic a oma ic hyd oca bons (PAH) and diesel
engine exhaus s, which may in e ac wi h he u o he-
lium o he bladde .7 Howe e , a case–con ol s udy om
he USA examined he li e ime occupa ional his o y o
Table 2 Bladde cance isk in men by occupa ion in NOCCA and CanCHEC
Occupa ional ca ego y
NOCCA (n=54 737) CanCHEC* (n=6630)
Cases SIR 95% CI Cases HR 95% CI
AM 2704 1.01 0.98 o 1.05 1115 1.11 1.04 o 1.19
AW 4356 0.74 0.72 o 0.77 455 0.84 0.76 o 0.93
CH 514 0.90 0.81 o 0.97 45 1.14 0.84 o 1.54
CS 53 1.47 1.10 o 1.92
CW 1704 1.06 1.01 o 1.10 410 0.95 0.86 o 1.06
DV 2899 0.99 0.95 o 1.03 520 0.95 0.87 o 1.04
EW 1519 1.05 1.00 o 1.11 150 0.88 0.75 o 1.04
FF 581 0.77 0.71 o 0.83 25 0.89 0.60 o 1.33
FB 64 1.09 0.84 o 1.39 160 1.00 0.85 o 1.17
FP 683 0.77 0.72 o 0.83 60 0.77 0.59 o 1.01
FW 762 0.80 0.75 o 0.86 85 1.02 0.82 o 1.27
HD 140 1.14 0.96 o 1.35 25 1.48 1.01 o 2.19
LW 80 1.04 0.82 o 1.29 5 0.46 0.22 o 0.97
MW 901 1.08 1.01 o 1.16 230 0.97 0.85 o 1.10
MN 246 1.06 0.93 o 1.20 60 1.12 0.87 o 1.45
PA 753 1.05 1.01 o 1.12 50 1.15 0.86 o 1.54
PL 469 1.02 0.93 o 1.12 90 0.95 0.77 o 1.16
PI 502 1.11 1.02 o 1.21 40 1.04 0.77 o 1.41
RP 713 1.04 0.97 o 1.12 30 1.22 0.84 o 1.78
SW 2514 1.11 1.07 o 1.15 525 1.12 1.02 o 1.22
TH 1372 0.84 0.80 o 0.90 230 0.87 0.76 o 1.00
TW 17 1.67 1.02 o 2.67
WA 122 1.41 1.17 o 1.68
WE 484 1.03 0.94 o 1.12 100 1.40 1.15 o 1.70
Emp y box indica es no in o ma ion a ailable. Bold indica es s a is ically signi ican inceased isk. I alics indica es s a is ically signi ican
dec eased isk.
*Model adjus ed o age, sex, egion and le el o educa ion.
AM, adminis a o s and manage s; AW, ag icul u e wo ke s; CanCHEC, Canadian Census Heal h and En i onmen Coho ; CH, chemical
wo ke s; CS, chimney-sweeps; CW, cons uc ion wo ke s; DV, d i e s; EW, elec ical wo ke s; FB, ood and be e age wo ke s; FF, i e
igh e s; FP, ood-p ocessing wo ke s; FW, o es y wo ke s; HD, hai d esse s; LW, laund y wo ke s; MN, mine s; MW, me al wo ke s; NOCCA;
No dic Occupa ional Cance s udy; PA, pain e s; PI, p in e s; PL, plumbe s; RP, ubbe and plas ics wo ke s; SIR, s anda dised incidence
a io; SW, sales wo ke s; TH, eache s; TW, obacco wo ke s; WA, wai e s; WE, welde s.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
6HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
pa icipan s24 and a na ionwide case–con ol s udy om
New Zealand25 did no obse ed he clea inc ease in isk
in hese s udies. The New Zealand s udy had obse ed
some e idence o inc eased isks o bladde cance
among emale d i e s.25 A me a-analysis26 sugges ed ha
he occupa ional isk o bladde cance among d i e s
has been educed in ecen yea s, which may be e lec ed
in ou indings o male d i e s.
This compa a i e s udy also obse ed e idence o
an associa ion be ween he isk o bladde cance
and employmen as a plumbe . The excess isk in he
CanCHEC coho was obse ed among a small numbe o
emale plumbe s in he s udy. Exis ing e idence has been
simila ly inconclusi e. S udies in he USA ha e epo ed
inc eased isks o bladde cance in male plumbe s24 27
bu his associa ion was no obse ed among plumbe s in
an I alian case–con ol s udy.28 Plumbe s a e exposed o
many haza dous ma e ials such as lead, welding umes,
a s and asbes os, which makes i di icul o iden i y he
pu a i e agen s.24 27
Pain e s a e commonly exposed o a oma ic amines
such as benzidine, 4-aminobiphenyl, β-naph hylamine
and 4-chlo o-o- oluidine and exposu e o hese subs ances
has been linked o bladde cance .29 The In e na ional
Agency o Resea ch on Cance (IARC) classi ied occu-
pa ional exposu e as a pain e as ca cinogenic o humans
(g oup 1) mainly based on obse ed inc eased isks o
lung cance and bladde cance . Al hough no eaching
s a is ical signi icance, pain e s appea ed o ha e an
inc eased isk o bladde cance in ou compa a i e s udy.
A me a-analysis epo ed a smoking-adjus ed signi ican
inc eased isk o bladde cance among pain e s, which
was consis en ly obse ed ac oss 36 obse a ional s udies
ega dless o s udy design, sex and s udy loca ion.30
Table 3 Bladde cance isk in women by occupa ion in he NOCCA and CanCHEC
Occupa ional ca ego ies
NOCCA (n=18 916) CanCHEC* (n=1540)
Cases SIR 95% CI Cases HR 95% CI
AM 137 1.17 1.00 o 1.40 110 0.93 0.77 o 1.13
AW 342 0.70 0.60 o 0.74 50 0.88 0.66 o 1.18
CH 63 1.45 1.11 o 1.86 <5 – –
CS 0 – 0.00 o 52.7
CW 11 1.00 0.50 o 1.80 10 1.25 0.63 o 2.51
DV 41 1.24 0.90 o 1.70 10 1.09 0.64 o 1.85
EW 72 1.00 0.79 o 1.26 5 0.65 0.27 o 1.57
FF 14 1.08 0.58 o 1.80 <5 – –
FB 12 1.07 0.55 o 1.87 125 1.07 0.89 o 1.29
FP 212 1.02 0.88 o 1.16 15 0.59 0.35 o 0.98
FW 4 0.65 0.18 o 1.65 <5 – –
HD 131 1.42 1.20 o 1.68 10 1.09 0.63 o 1.89
LW 131 1.18 0.98 o 1.40 15 1.11 0.65 o 1.88
MW 10 0.92 0.44 o 1.70 10 1.44 0.72 o 2.88
MN 2 1.75 0.21 o 6.34 <5 – –
PA 8 1.29 0.56 o 2.54 <5 – –
PL 0 – 0.00 o 21.7 5 2.90 1.20 o 6.98
PI 72 1.47 1.15 o 1.85 10 1.21 0.63 o 2.34
RP 101 1.01 0.82 o 1.22 <5 – –
SW 205 1.27 1.10 o 1.46 185 1.06 0.91 o 1.23
TH 458 0.81 0.73 o 0.90 125 0.91 0.75 o 1.11
TW 22 1.65 1.03 o 2.50
WA 268 1.30 1.15 o 1.47
WE 4 1.08 0.30 o 2.77 <5 – –
*Model adjus ed o age, sex, egion and le el o educa ion.
Bold indica es s a is ically signi ican inc eased isk. I alics indica e s a is ically signi ican dec eased isk.
AM, adminis a o s and manage s; AW, ag icul u e wo ke s; CanCHEC, Canadian Census Heal h and En i onmen Coho ; CH, chemical
wo ke s; CS, chimney-sweeps; CW, cons uc ion wo ke s; DV, d i e s; EW, elec ical wo ke s; FB, ood and be e age wo ke s; FF, i e
igh e s; FP, ood-p ocessing wo ke s; FW, o es y wo ke s; HD, hai d esse s; LW, laund y wo ke s; MN, mine s; MW, me al wo ke s; NOCCA;
No dic Occupa ional Cance s udy; PA, pain e s; PI, p in e s; PL, plumbe s; RP, ubbe and plas ics wo ke s; SIR, s anda dised incidence
a io; SW, sales wo ke s; TH, eache s; TW, obacco wo ke s; WA, wai e s; WE, welde s.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
7
HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
Though he speci ic agen s a ibu able o he isk o
bladde cance in pain e s ha e no been clea ly iden i-
ied, exposu e o me al coa ings, wood a nishes o s ains
ha con ain sol en s, ille s and pigmen s a e sugges ed
isk ac o s.30
Mine s appea ed o be a inc eased isk o bladde
cance in ou compa a i e s udy. This associa ion was
p e iously epo ed in s udies om bo h Eu ope8 20 21 and
Canada.8 31 Mine s ha e known occupa ional exposu es
o chemical compounds such as PAH, which ha e been
sugges ed as a s ong isk ac o in he ae iology o bladde
cance .32 This inc eased isk could be due o a numbe o
exposu es in he mining se ing. Mine s can be exposed
o ela i ely high le els o diesel exhaus , p ima ily among
unde g ound anspo a ion wo ke s, whe e excess isks
o bladde cance ha e p e iously been obse ed.31
Inc eased isk was obse ed among emale d y cleane s
in ou compa a i e s udy, al hough isk es ima es we e
no s a is ically signi ican . A ecen me a-analysis o d y
cleaning wo ke s epo ed a signi ican ly ele a ed isk
o bladde cance , and he excess isk did no appea o
be con ounded by smoking.33 A case–con ol s udy om
he USA also obse ed a simila inc eased isk o bladde
cance among d y cleane s, bu his s udy was limi ed o
non-whi e men.9 In he NOCCA s udy, i was no possible
o sepa a e d y cleane s om o he laund y wo ke s
and hence we used d y cleane s as a compa ison g oup
o launde e s. Use o e achlo oe hylene was by a he
dominan sol en among No dic d y cleane s. Laun-
de e s can be exposed o e achlo oe hylene, which is a
po en ial bladde ca cinogen acco ding o IARC (g oup
2A).33
Ou s udy obse ed signi ican ly ele a ed isk o bladde
cance associa ed wi h welding only among men in he
CanCHEC coho . An IARC wo king g oup had p e i-
ously explo ed his associa ion, and i was deemed limi ed
and inconclusi e.34 Since he 1990 IARC e alua ion, many
s udies ha e epo ed ele a ed isks o bladde cance
associa ed wi h welding. Se e al o hese s udies ha e,
howe e , included small sample sizes o small numbe s
o cases. A hospi al-based case–con ol s udy in F ance
epo ed a se en old inc eased isk o bladde cance
in male welde s compa ed wi h con ols.32 Howe e ,
he associa ion obse ed was no s a is ically signi ican .
Excess isk o bladde cance was also epo ed in he
Ge man-based s udy among welde s who had wo ked o
a leas 6 mon hs in his occupa ion a he ime o i s
ollow-up (s anda dised mo ali y a io 3.04, 95% CI 1.14
o 8.10), bu he isk es ima e was obse ed o be non-s a-
is ically signi ican wi h ex ended ollow-up.35 While
he s onges e idence o cance isk in welde s is lung
cance , he s udy sugges s ha occupa ional associa ions
wi h bladde cance a e mos p obably due o exposu es
speci ic o welding umes ha may con ibu e o he isk
o bladde cance .12
Reduced isks o bladde cance we e associa ed wi h
employmen as a eache and ag icul u al wo ke . These
p o ec i e e ec s ha e been epo ed in a p e ious
me a-analysis.36 Consis en ly, lowe isk o bladde cance
was obse ed among ag icul u al wo ke s in Spanish
s udies.20 21 Low p e alence o smoking, high physical
demand o exe cise in a ming occupa ions, die a y ac o s
such as high in ake o esh ood and ege ables and esi-
dence in a eas wi h li le ai pollu ion migh accoun o
he educed isk o bladde cance among a me s.21 36 In
con as , a 2009 s udy om he USA epo ed long- e m
employmen as an ag icul u al wo ke was associa ed wi h
inc eased isk o bladde cance .37 The s udy sugges ed
ha long- e m exposu e o ca cinogenic agen s such
as pes icides, sol en s and o he ino ganic dus s could
con ibu e o he inc eased isk o bladde cance among
ag icul u al wo ke s.37
Smoking is he s onges known isk ac o o bladde
cance and an impo an po en ial con ounde in he
s udies o occupa ional bladde cance .38 The CanCHEC
s udy allowed o adjus men o educa ion, which
aimed o indi ec ly con ol o occupa ional di e ence
in smoking. Smoking has been shown o be s ongly
co ela ed wi h educa ion le els and has been sugges ed
as an app op ia e p oxy in model adjus men in he
absence o indi idual-le el smoking da a.39 The s udy also
obse ed highe isk o smoking associa ed wi h lowe
educa ion le el. In ou NOCCA–CanCHEC compa a i e
s udy, hose occupa ional g oups a highes isk o bladde
cance we e also commonly iden i ied as a isk o lung
cance . Co ela ion o bladde and lung cance in bo h
he NOCCA and he CanCHEC s udies in men sugges s
ha smoking o some o he sha ed isk ac o (s) may
be esponsible o excess isks obse ed by occupa ions
( igu e 2). Howe e , we also obse ed some ou lie s in he
g aph including d y cleane s, hai d esse s and welde s,
which may indica e some occupa ion-speci ic isk ac o s
o hese g oups.
P e ious a emp s o disen angle he e ec s o occu-
pa ional exposu es om smoking ha e indica ed ha
smoking does no accoun o all o he excess isk o
bladde cance .12 40 Al hough hai d esse s ha e a highe
p e alence o smoking compa ed wi h o he wo ke s and
he gene al popula ion,16 Takkouche e al obse ed no
di e ence in e ec among his g oup wi h adjus men o
smoking.15 Adjus men o smoking also appea ed o ha e
li le o no e ec on isk es ima es o pain e s,30 hai -
d esse s,14 p in e s, anspo a ion wo ke s,20 machine
ope a o s21 and plumbe s.27 This suppo s he hypo h-
esis ha a leas some occupa ional a ia ion in bladde
cance isk can be explained by occupa ional di e ence
in smoking.12 Howe e , we canno ule ou he possibili y
o esidual con ounding in ou s udy.
This compa ison o wo la ge coho s udies conduc ed
in di e en geog aphic egions aimed o desc ibe a ia-
ion in bladde cance isk ac oss occupa ions was unable
o iden i y speci ic occupa ional pa e ns in isk o bladde
cance excep o sales wo ke s. Fu he mo e, No dic
peoples ha e long occupa ional his o y as compa ed wi h
he Canadians. I we we e unable o iden i y he speci ic
isk o bladde cance ha would be simila ly high in bo h
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
8HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
geog aphic egions, one o he easons may be ha some
o he isk es ima es in one o bo h egions may be dilu ed
owa ds uni y. I only a small ac ion o he wo ke s a e
exposed, he e ec may no be seen in he isk es ima e
o he en i e occupa ional ca ego y. Addi ionally, occupa-
ion a one poin in ime does no necessa ily co espond
o li elong occupa ional his o y. Some o he pe sons in
non-exposed ca ego ies may ac ually ha e had exposu e
o bladde ca cinogens in o he jobs. This would esul
in misclassi ica ion and an a enua ion o isk es ima es.
S eng hs o his s udy include he la ge coho sizes, long
du a ion o ollow-up and la ge numbe o cases obse ed,
al hough he NOCCA coho was s onge in all o hese
aspec s. Howe e , he numbe s o cases in some o he
smalle occupa ional ca ego ies we e oo ew o p ecisely
measu e associa ions. The use o exis ing egis ies cap u ed
i ually all inciden cance cases. The linkage be ween he
census, mo ali y and immig a ion, and cance incidence
da a we e based on unique pe sonal iden i y codes used in
egis ies in all No dic coun ies12 and h ough de e min-
is ic and p obabilis ic me hods in CanCHEC ha ensu es
a high p obabili y o asce aining o ele an e en s. The
use o exis ing da a sou ces including he censuses and
mo ali y and mobili y eco ds was an e icien app oach o
su eillance o la ge popula ions.
COnClusIOn
We obse ed di e en occupa ional pa e ns in isk o
bladde cance in No dic coun ies and Canada. This
compa a i e s udy iden i ied consis ency in isk only
among sales wo ke s. Risks o bladde cance a ied ac oss
occupa ional ca ego ies. Al hough exposu e o ca cino-
gens a wo k may con ibu e his a ia ion, howe e , he
p esen compa ison o bladde cance pa e n in he
No dic coun ies and in Canada did no iden i y clues
o he disease ae iology. The s udy illus a es ha he
possible e ec o speci ic occupa ional exposu e may be
di icul o un a el in he da ase s ollowing people wi h
wo k expe iences om di e en ime pe iods and ca ego-
ies in o b oad occupa ional g oups.
Au ho a ilia ions
1Depa men o Epidemiology, Facul y o Social Sciences, Uni e si y o Tampe e,
Tampe e, Finland
2Occupa ional Cance Resea ch Cen e, To on o, Canada
3Dalla Lana School o Public Heal h, Uni e si y o To on o, To on o, On a io, Canada
4CAREX Canada, Facul y o Heal h Sciences, Simon F ase Uni e si y, Bu naby,
Canada
5Depa men o Resea ch, Cance Regis y o No way, Ins i u e o Popula ion-Based
Cance Resea ch, Oslo, No way
6Depa men o Communi y Medicine, Facul y o Heal h Sciences, Uni e si y o
T omsø, The A c ic Uni e si y o No way, T omsø, No way
7Gene ic Epidemiology G oup, Folkhälsan Resea ch Cen e , Helsinki, Finland
8Depa men o Medical Epidemiology and Bios a is ics, Ka olinska Ins i u e ,
S ockholm, Sweden
9Cen e o Epidemiology and Sc eening, Ins i u e o Public Heal h, Uni e si y o
Copenhagen, Copenhagen, Denma k
10Icelandic Cance Regis y, Reykja ik, Iceland
11Facul y o Medicine, Uni e si y o Iceland, Reykja ik, Iceland
12School o Occupa ional and Public Heal h, Rye son Uni e si y, To on o, Canada
13Heal h Analysis Di ision, S a is ics Canada, O awa, On a io, Canada
14Depa men o Sociology and Economics, Uni e si y o New B unswick F ede ic on
Campus, New B unswick, Canada
15Finnish Cance Regis y, Ins i u e o S a is ical and Epidemiological Cance
Resea ch, Helsinki, Finland
Con ibu o s KH, JM, PAD and EP designed, analysed and p epa ed he
manusc ip . KH, JIM, EW, KK, EL, PS, LT and EP a e esponsible o NOCCA and JM
and PAD a e esponsible o CanCHEC coho s. KH, JM, PAD, JIM, EW, KK, EL, PS, LT,
MAH, MT, PAP and EP e ised, e iewed, p o ided c i ical eedbacks and app o ed
he manusc ip o inal submission.
Funding This wo k was suppo ed by unding om he No dic Cance Union. The
CanCHEC analysis was unded by a g an om he On a io Wo kplace Sa e y and
Insu ance Boa d Resea ch Ad iso y Council (#11024) and suppo ed by he On a io
Minis y o Labou and he Canadian Cance Socie y Resea ch Ins i u e. Linkage
o he 1991 Canadian Census o he Canadian Cance Regis y was unded by he
Canadian Ins i u e o Heal h In o ma ion, Heal h Canada and S a is ics Canada. This
esea ch was suppo ed by unds o he Canadian Resea ch Da a Cen e Ne wo k
om he Social Science and Humani ies Resea ch Council, he Canadian Ins i u e
o Heal h Resea ch, he Canadian Founda ion o Inno a ion and S a is ics Canada.
Al hough he esea ch and analysis a e based on da a om S a is ics Canada, he
opinions exp essed do no ep esen he iews o S a is ics Canada o he Canadian
Resea ch Da a Cen e Ne wo k.
Compe ing in e es s None decla ed.
e hics app o al P io app o al was ob ained and no addi ional app o al was
needed o his speci ic s udy.
P o enance and pee e iew No commissioned; ex e nally pee e iewed.
da a sha ing s a emen No addi ional da a a e a ailable.
Open Access This is an Open Access a icle dis ibu ed in acco dance wi h he
C ea i e Commons A ibu ion Non Comme cial (CC BY-NC 4.0) license, which
pe mi s o he s o dis ibu e, emix, adap , build upon his wo k non-comme cially,
and license hei de i a i e wo ks on di e en e ms, p o ided he o iginal wo k is
p ope ly ci ed and he use is non-comme cial. See: h p:// c ea i ecommons. o g/
licenses/ by- nc/ 4. 0/
© A icle au ho (s) (o hei employe (s) unless o he wise s a ed in he ex o he
a icle) 2017. All igh s ese ed. No comme cial use is pe mi ed unless o he wise
exp essly g an ed.
eFe enCes
1. Vineis P, Simona o L. P opo ion o lung and bladde cance s in
males esul ing om occupa ion: a sys ema ic app oach. A ch
En i on Heal h 1991;46:6–15.
2. D eye L, Win he JF, Pukkala E, e al. A oidable cance s in he
No dic coun ies. Tobacco smoking. APMIS Suppl 1997;76:9–47.
3. In e na ional Agency o esea ch on Cance . GLOBOCAN:
Es ima ed Cance Incidence, Mo ali y and P e alence wo ldwide
in 2012. h p:// globocan. ia c. / Pages/ ac _ shee s_ popula ion. aspx.
(accessed 20 Dec 2015).
4. Associa ion o he No dic Cance Regis ies. NORDCAN: cance
incidence and mo ali y in he no dic coun ies, Ve sion 7.1. 2015.
h p://www. anc . nu. (accessed 18 Aug 2015).
5. Canadian Cance s a is ics 2015. Special opic: P edic ions o he
u u e bu den o cance in Canada. h ps://www. cance . ca/~/ media/
cance . ca/ CW/ cance % 20in o ma ion/ cance % 20101/ Canadian%
20cance % 20s a is ics/ Canadian- Cance - S a is ics- 2015- EN. pd .
(accessed 28 Jan 2016).
6. Cogliano VJ, Baan R, S ai K, e al. P e en able exposu es
associa ed wi h human cance s. J Na l Cance Ins
2011;103:1827–39.
7. Sil e man DT, Hoo e RN, Mason TJ, e al. Mo o exhaus - ela ed
occupa ions and bladde cance . Cance Res 1986;46(4 P
2):2113–6.
8. Reulen RC, Kellen E, Bun inx F, e al. A me a-analysis on he
associa ion be ween bladde cance and occupa ion. Scand J U ol
Neph ol Suppl 2008;218:64–78.
9. Sil e man DT, Le in LI, Hoo e RN. Occupa ional isks o bladde
cance in he Uni ed S a es: II Nonwhi e men. J Na l Cance Ins
1989;81:1480–3.
10. Ji J, G ans öm C, Hemminki K. Occupa ion and bladde cance : a
coho s udy in Sweden. B J Cance 2005;92:1276–8.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om
9
HadkhaleK, e al. BMJ Open 2017;7:e016538. doi:10.1136/bmjopen-2017-016538
Open Access
11. Engholm G, Hakulinen T, Gislum M, e al. T ends in he su i al o
pa ien s diagnosed wi h kidney o u ina y bladde cance in he
No dic coun ies 1964-2003 ollowed up o he end o 2006. Ac a
Oncol 2010;49:655–64.
12. Pukkala E, Ma insen JI, Lynge E, e al. Occupa ion and cance -
ollow-up o 15 million people in i e No dic coun ies. Ac a Oncol
2009;48:646–790.
13. Pe e s PA, Tjepkema M, Wilkins R, e al. Da a esou ce p o ile: 1991
Canadian Census Coho . In J Epidemiol 2013;42:1319–26.
14. Ha ling M, Schablon A, Schedlbaue G, e al. Bladde cance among
hai d esse s: a me a-analysis. Occup En i on Med 2010;67:351–8.
15. Takkouche B, Reguei a-Méndez C, Mon es-Ma ínez A. Risk o
cance among hai d esse s and ela ed wo ke s: a me a-analysis. In
J Epidemiol 2009;38:1512–31.
16. IARC monog aphs on he e alua ion o ca cinogenic isks o humans.
occupa ional exposu e o hai d esse s and ba be s and pe sonal use
o hai colo an s; some hai dyes; cosme ic colo an s, indus ial dye
s u s and a oma ic amines. IARC Monog E al Ca cinog Risk Hum
1993;57:1–437 h p:// monog aphs. ia c. / ENG/ Monog aphs/ ol57/
mono57. pd .
17. Fe ís J, Ga cia J, Be bel O, e al. Cons i u ional and occupa ional
isk ac o s associa ed wi h bladde cance . Ac as U ol Esp
2013;37:513–22.
18. Pukkala E, Nokso-Koi is o P, Roponen P. Changing cance isk
pa e n among Finnish hai d esse s. In A ch Occup En i on Heal h
1992;64:39–42.
19. Bol HM, Golka K. The deba e on ca cinogenici y o pe manen hai
dyes: new insigh s. C i Re Toxicol 2007;37:521–36.
20. Koge inas M, ' Manne je A, Co die S, e al. Occupa ion and bladde
cance among men in Wes e n Eu ope. Cance Causes Con ol
2003;14:907–14.
21. Samanic CM, Koge inas M, Sil e man DT, e al. Occupa ion and
bladde cance in a hospi al-based case-con ol s udy in Spain.
Occup En i on Med 2008;65:347–53. 2008.
22. IARC monog aphs on he e alua ion o ca cinogenic isks o
humans. p in ing p ocesses and p in ing inks, ca bon black and
some ni o compounds. IARC Monog E al Ca cinog Risk Hum
1996;65:1–115 h p:// monog aphs. ia c. / ENG/ Monog aphs/ ol65/
mono65- 5. pd .
23. ' Manne je A, Pea ce N. Bladde cance isk in sales wo ke s:
a e ac o cause o conce n? Am J Ind Med 2006;49:175–86.
24. Col JS, Ka agas MR, Schwenn M, e al. Occupa ion and bladde
cance in a popula ion-based case-con ol s udy in No he n New
England. Occup En i on Med 2011;68:239–49.
25. D yson E, ' Manne je A, Walls C, e al. Case-con ol s udy o high
isk occupa ions o bladde cance in New Zealand. In J Cance
2008;122:1340–6.
26. Manju L, Geo ge PS, Ma hew A. U ina y bladde cance isk among
mo o ehicle d i e s: a me a-analysis o he e idence, 1977-2008.
Asian Pac J Cance P e 2009;10:287–94.
27. Zheng T, Can o KP, Zhang Y, e al. Occupa ion and bladde cance :
a popula ion-based, case-con ol s udy in Iowa. J Occup En i on
Med 2002;44:685–91.
28. Pun oni R, Me lo F, Bo sa L, e al. A his o ical coho mo ali y
s udy among shipya d wo ke s in Genoa, I aly. Am J Ind Med
2001;40:363–70.
29. Bachand A, Mund KA, Mund DJ, e al. Me a-analyses o
occupa ional exposu e as a pain e and lung and bladde cance
mo bidi y and mo ali y 1950-2008. C i Re Toxicol 2010;40:101–25.
30. Guha N, S eenland NK, Me le i F, e al. Bladde cance isk in
pain e s: a me a-analysis. Occup En i on Med 2010;67:568–73.
31. Gae ne RR, T peski L, Johnson KC.A case-con ol s udy o
occupa ional isk ac o s o bladde cance in Canada. Cance
Causes Con ol 2004;15:1007–19.
32. Co die S, Cla el J, Limasse JC, e al. Occupa ional isks o bladde
cance in F ance: a mul icen e case-con ol s udy. In J Epidemiol
1993;22:403–11.
33. Vlaande en J, S ai K, Rude A, e al. Te achlo oe hylene exposu e
and bladde cance isk: a me a-analysis o d y-cleaning-wo ke
s udies. En i on Heal h Pe spec 2014;122:661–6.
34. IARC monog aphs on he e alua ion o ca cinogenic isks o humans.
Ch omium, Nickel and welding. IARC Monog E al Ca cinog Risk
Hum 1990;49:1–687 h p:// monog aphs. ia c. / ENG/ Monog aphs/
ol49/ mono49. pd .
35. Becke N. Cance mo ali y among a c welde s exposed o umes
con aining ch omium and nickel. Resul s o a hi d ollow-up: 1989-
1995. J Occup En i on Med 1999;41:294–303.
36. Blai A, Zahm SH, Pea ce NE, e al. Clues o cance e iology om
s udies o a me s. Scand J Wo k En i on Heal h 1992;18:209–15.
37. Cassidy A, Wang W, Wu X, e al. Risk o u ina y bladde cance :
a case-con ol analysis o indus y and occupa ion. BMC Cance
2009;9:443.
38. In e na ional Agency o Resea ch on Cance . IARC monog aphs on
he e alua ion o ca cinogenic isk o humans. Agen s classi ied by
he IARC monog aphs, Volumes 1 o 117 h p:// monog aphs. ia c. /
ENG/ Classi ica ion/ index. php.
39. Winkleby MA, Ja ulis DE, F ank E, e al. Socioeconomic s a us
and heal h: how educa ion, income, and occupa ion con ibu e
o isk ac o s o ca dio ascula disease. Am J Public Heal h
1992;82:816–20.
40. Shiels MS, Gibson T, Sampson J, e al. Ciga e e smoking p io
o i s cance and isk o second smoking-associa ed cance s
among su i o s o bladde , kidney, head and neck, and s age I lung
cance s. J Clin Oncol 2014;32:3989–95.
g oup.bmj.com on Sep embe 5, 2017 - Published by h p://bmjopen.bmj.com/Downloaded om