E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

. sc iencedi rec t . com

ava i lab le a t www

journal homepage: www.ejconl ine.com

Cancer risk diversity in non-western migrants to Europe: Anoverview of the literature

Melina Arnold a,b,*, Oliver Razum a, Jan-Willem Coebergh b,c

a Department of Epidemiology and International Public Health, Bielefeld University, Germanyb Department of Public Health, Erasmus Medical Centre Rotterdam, The Netherlandsc Comprehensive Cancer Centre South, Eindhoven Cancer Registry (IKZ), The Netherlands

A R T I C L E I N F O

Article history:

Received 4 January 2010

Received in revised form 11 May

2010

Accepted 29 July 2010

Keywords:

Migrants

Cancer

Inequalities

Health transition

Europe

Review literature

0959-8049/$ - see front matter � 2010 Elsevidoi:10.1016/j.ejca.2010.07.050

* Corresponding author: Address: Bielefeld UHealth, University of Bielefeld, P.O. 10 01 31,

E-mail address: melina.arnold@uni-bielef

A B S T R A C T

Background: Cancer risk varies geographically and across ethnic groups that can be moni-

tored in cancer control to respond to observed trends as well as ensure appropriate health

care. The study of cancer risk in immigrant populations has great potential to contribute

new insights into aetiology, diagnosis and treatment of cancer. Disparities in cancer risk

patterns between immigrant and autochthonous populations have been reported many

times, but up to now studies have been heterogeneous and may be discordant in their find-

ings. The aim of this overview was to compile and compare studies on cancer occurrence in

migrant populations from non-western countries residing in Western Europe in order to

reflect current knowledge in this field and to appeal for further research and culturally sen-

sitive prevention strategies.

Methods: We included 37 studies published in the English language between 1990 and April

2010 focussing on cancer in adult migrants from non-western countries, living in the

industrialised countries of the European Union. Migrants were defined based on their coun-

try of birth, ethnicity and name-based approaches. We conducted a between-country com-

parison of age-adjusted cancer incidence and mortality in immigrant populations with

those in autochthonous populations.

Findings: Across the board migrants from non-western countries showed a more favourable

all-cancer morbidity and mortality compared with native populations of European host

countries, but with considerable site-specific risk diversity: Migrants from non-western

countries were more prone to cancers that are related to infections experienced in early

life, such as liver, cervical and stomach cancer. In contrast, migrants of non-western origin

were less likely to suffer from cancers related to a western lifestyle, e.g. colorectal, breast

and prostate cancer.

Discussion: Confirming the great cancer risk diversity in non-western migrants in and

between different European countries, this overview reaffirms the importance of exposures

experienced during life course (before, during and after migration) for carcinogenesis. Cul-

turally sensitive cancer prevention programmes should focus on individual risk patterns

and specific health care needs. Therefore, continuously changing environments and subse-

quently changing risks in both migrant and autochthonous populations need to be

observed carefully in the future.

� 2010 Elsevier Ltd. All rights reserved.

er Ltd. All rights reserved.

niversity, School of Public Health, Department of Epidemiology and International PublicD-33501 Bielefeld, Germany. Tel.: +49 (0)521 106 2539; fax: +49 (0)521 106 6465.eld.de (M. Arnold).

2648 E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

1. Backgroundsults from studies conducted all over Europe dealing with

Studies on cancer risk in migrant populations have recently

gained increased recognition, but still have rather heteroge-

neous study populations and methods applied. However, in-

sights into risk diversity deduced from such studies

contribute to our understanding of carcinogenesis and might

help answer unclear aetiology questions.

Migration has become an important phenomenon in Wes-

tern Europe in terms of population changes and the composi-

tion of society during the past decades. In 2005, Western and

Central Europe hosted 44.1 million migrants, defined as for-

eign-born persons.1 Many of them originate from non-wes-

tern countries, seeking social security, employment

opportunities and a better future.

European societies characterised by an increasing degree

of heterogeneity pose major challenges to health care sys-

tems and policies. Evidence-based research is therefore a pre-

requisite for appropriate and individual health care of high

quality and effectiveness as well as the implementation of

culturally sensitive measures of prevention.2,3

Health is closely related to global movements. The transi-

tion of disease and risk patterns over time and across coun-

tries have been the scope of many epidemiological research

questions. Accordingly, infectious diseases become less

important as populations advance in terms of westernisation

and the role of chronic health conditions, such as cancer and

cardio-vascular diseases, becomes predominant.4

Hence, migrants from non-western countries are equipped

with a unique constellation of risk factors that are deter-

mined by exposure and disease patterns experienced in both

their home as well as their host country.5,6 This sudden

change in the stage of epidemiological transition as well as

environmental determinants has a major impact on an indi-

vidual’s lifetime disease risk.

Many theories have been developed to explain differences

in mortality and morbidity between migrants and the popula-

tion of their host and home countries, respectively, one of them

being the healthy migrant effect. Thus, migrants are subject to

selection processes that initially underlie good physical and

mental health. Those health advantages after migration are

thought likely to disappear with advancing duration of resi-

dence and generations. As suggested in some studies, no evi-

dence of quickly diminishing health advantages could be

observed, challenging this concept and allowing room for other

explanations.7 Nonetheless, the change in risk patterns over

time is of special interest in epidemiological research.

Multi-causality and geographical variation make cancer in

migrant populations highly suitable for research, especially

in cancers whose main causes are still not attributable to either

environmental (‘nurturecomponents’) orgenetic (‘nature com-

ponents’) risk factors.8 In this context, the individual life course

and particularly early life experiences (as the first step in carci-

nogenesis) have a great impact and play a major role in the

effects of exposure and their association with cancer risks.9,10

Investigating the occurrence of cancer in migrant popula-

tions may allow for a better understanding of cancer aetiology

and of biological factors that can be integrated into preven-

tion and treatment programmes.

The purpose of this article is to compile and compare re-

cancer in non-western migrant populations. The resulting

overview can serve as a guide, reflecting the present state of

knowledge in this field, and as an appeal for further research

and prevention.

2. Methods

2.1. Inclusion criteria of studies

We included studies focussing mainly or partly on cancer

incidence and mortality in adult migrants from non-western

countries, living in the industrialised countries of the Euro-

pean Union, published in English between 1990 and April

2010. Studies were identified by searching pubmed and other

established scientific databases in combination with the fol-

lowing keywords: cancer + ethnicity/ethnic minority/(im)mi-

grant(s)/foreign(ers)/country of birth. A further inclusion

criterion was a comparison of the migrant population with

the native population of the country of the study (no studies

conducted within migrant populations).

2.2. Study descriptions

We identified 37 studies conducted in the following seven

countries: Denmark (3), France (4), Germany (6), Spain (1),

Sweden (7), The Netherlands (5) and the United Kingdom

(11). In 51% of the studies (19/37) incidence data were ana-

lysed, in 41% (15/37) mortality data and in 8% (3/37) both.

All studies were based on the retrospective cohort design.

Owing to the heterogeneous measures of association ap-

plied in the studies, we described tendencies instead of com-

bined rate ratios (RRs) or odds ratios (ORs) to indicate

differences in risks as follows: significantly elevated, elevated,

no difference, decreased and significantly decreased. Age-

adjustment procedures had been carried out in all the studies

included. Other covariables are listed in Table 1.

In general 70% of the studies (26/37) involved all-cancer

comparisons and 24% of the studies (9/37) focused on only

one specific cancer site. The most commonly investigated

sites were breast (28 studies) and lung cancer (26 studies) as

well as stomach and colorectal cancer (24 studies each).

2.3. Defining the migrant status, generations involvedand pooling of migrant origins

The indicator for defining the migrant population under study

ranged from country of birth (of the patient or in combination

with the parental country of birth) in 73% (27/37), name-based

approaches in 14% (5/37), (self-assigned) ethnicity in 11% (4/

37) and a combination in one study.

The applied indicator or proxy for ethnicity is highly

dependent on the availability and completeness of potential

variables in the particular host country. However, country of

birth is the most widely used and accepted proxy although

it has some validity limitations with regard to cultural and

ethnic identity.11

Table 1 – Methodological features of the studies included.

Country, authors and year

of study

Study aim: to explore Data source Period Outcome/measure of

association (covariables)

Cohort acquisition/In- and

exclusion criteria

Methodological peculiarities Definition of ethnicity Reference population Size and composition of

study population

Discussed explanations

for risk differences

Study limitations

Denmark Myrup et al.

(2008)13

The aetiology of

testicular cancer risk

Study population: civil

registration system linked

to Danish Cancer Registry

through unique personal

identification number

(population-based)

1968–2003 Incidence RR (Age,

calendar year, parental

birthplace, duration of stay,

age at immigration)

Males born between 1930 and

2003; residents of Denmark

between 2nd April 1968 and 31st

December 2003, born between 1st

January 1930 and 31st December

2003; known place of birth;

exclusion of individuals born in

Greenland

Adjustments for maternal and

paternal birthplace in different

strata; trend analyses for duration of

stay and age at immigration

(Parental) Country of birth

(collected by civil

registration system from

index cards in

municipality registration

offices)

Men born in Denmark of

parents born in

Denmark

Cohort: n = 2,109,459

Cancer cases in cohort:

n = 4216 (1st generation

migrants: n = 166 (3.9%),

2nd generation

migrants: n = 13 (0.3%)

Early environmental

exposures/period in uteri;

salmon bias

Small number of cases

in second-generation

immigrants

Norredam

et al. (2008)29

Differences in cancer

stage at diagnosis

between migrant

women and native

Danish women

Study population:

Statistical Department at

The Danish Immigration

Service; linkage of civil

registration numbers of

the study cohort with

Danish Cancer Registry

(population-/register-

based cohort)

1993–1999

(cohort)/2002

(cases)

Incidence OR (matching

procedure, age group,

cancer type at first

diagnosis)

Women aged 18+; migrants with

residence permit as refugees or

through family reunification in

Denmark between 1st January

1993 and 31st December 1999;

only first diagnosis cancers; only

cancer types allowing

categorisation of stage; exclusion

criteria: missing civil registration

number; duplicates; unclear or

missing data on nationality

1:6 matching on age and sex on

population level; 1:4 matching on an

individual level on age and sex

through a random sampling

procedure; comparison of local with

non-local stages of tumours;

migrant status as proxy for pre- and

post-migration circumstances

Nationality according to

WHO’s classification

system

Danish-born residents

with Danish-born

parents (identified

through Statistics DK)

Study cohort: Cases (1st

generation Migrants)

n = 62461; Controls

(Danish-born)

n = 249,839; Cancer

cohort: Cases n = 269;

Controls n = 1608

Differences in tumour

biology between migrants

and host populations;

barriers in access to

healthcare (language,

culture, health care

system); poor screening

uptake; salmon bias

Small number of cases;

high number of cases

with unknown stage;

nationality as poor bio-

socio-cultural proxy of

ethnicity; no SES

adjustments possible

Norredam

et al. (2007)30

Incidence of cancer

among 1st generation

migrants compared

with native Danes,

including time trends

Statistical Department at

The Danish Immigration

Service; linkage of civil

registration numbers of

the study cohort with

Danish Cancer Registry

(population-/register-

based cohort)

1993–2003 Incidence RR (age, region

of origin, migrant type,

duration of residence)

Men and women aged 30–80;

residence permit as refugees or

through family reunification in

Denmark between 1st January

1993 and 31st December 1999;

exclusion criteria: missing civil

registration number; duplicates;

unclear or missing data on

nationality; non-melanoma skin

cancers

1:6 matching on age and sex upon

arrival in Denmark and 1:4

matching on an individual level on

age and sex through a random

sampling procedure in the study

cohort; migrant status as proxy for

pre- and post-migration

circumstances

Nationality according to

WHO’s classification

system

Danish-born residents

with Danish-born

parents (identified

through Statistics DK)

Study cohort: cases (1st

generation migrants)

n = 62461; controls

(Danish-born)

n = 249,899; cancer cases

n = 3366 (16% migrants)

Lifestyle patterns (breast

and colorectal cancer),

smoking; decline in

incidence over time in

migrant women related to

increased cancer

diagnostic activities such

as screening and better

access to healthcare

services

Small number of cases;

no SES adjustments

possible; trend analysis

irrespective of duration

of stay which may dilute

effects

France Bouchardy

et al. (1996)31

Cancer mortality in

North African

migrants to France

Population data: ‘Institut

National de la Statistique

et des Etudes

Economiques’ (INSEE),

derived from the French

1982 census; mortality

data: ‘Institut National de

la sante et de la recherche

medicale’ (INSERM)

1979–1985 Mortality RR (age, gender,

social class, area of

residence)

Men and women of all ages;

records of deaths in resident

population of France from 1979 to

1985 (provided by INSERM)

Stratified analyses by

socioeconomic subgroup

Country of birth Individuals born in

metropolitan France

(native French)

Cancer deaths among

migrants: n = 27,352

(3.4% of all cancer

deaths)

Return of ill migrants to

country of origin prior to

death; healthy-migrant

effect; lower consumption

of alcohol and higher

tobacco intake; dietary

differences; reproductive

behaviour; cultural

factors related to Islam

Poor quality of French

mortality data; small

numbers of cancer

deaths among Egyptian

migrants; heterogeneity

within migrant groups

Bouchardy

et al. (1995)32

Cancer mortality in

sub-Saharan African

migrants to France

Population data: ‘Institut

National de la Statistique

et des Etudes

Economiques’ (INSEE),

derived from the French

1982 census; mortality

data: ‘Institut National de

la sante et de la recherche

medicale’ (INSERM)

1979–1985 Mortality RR (age group,

gender, social class, area

of residence)

Men and women of all ages;

records of deaths in resident

population of France from 1979 to

1985 (provided by INSERM)

Stratified analyses by

socioeconomic subgroup

Country of birth Individuals born in

metropolitan France

(native French)

Migrant study

population: n = 288,060;

Cancer deaths among

migrants: n = 1126 (0.1%)

Return of ill migrants to

country of origin; healthy

migrant effect; protective

lifestyle factors (tobacco,

alcohol consumption,

lower meat/fat intake,

high fibre diets); infection

with hepatitis B virus

during childhood/chronic

persistent hepatitis (liver

cancer); Schistosoma

haematobium infections

(bladder cancer); Burkitt’s

lymphoma (NHL)

Poor data quality;

heterogeneity within

migrant groups

Khlat

(1995)33

The cancer profile of

Maghrebian and Near

Eastern migrants

Two large migrant studies 1979–1991 Mortality RR (age, area of

residence, social class)

Men and women of all ages;

French mortality data

Review of studies Country of birth Native French

population

Cancer deaths among

Moroccan migrants:

n = 2062

Genetic factors, diet,

alcohol consumption,

childbearing patterns,

cultural factors, viral

causes

Bouchardy

et al. (1994)34

cancer patterns in

Chinese and South

East Asian migrants

to France

Population data: « Institut

National de la Statistique

et des Etudes

Economiques » (INSEE),

derived from the French

1982 census; mortality

data: « Institut National

de la sante et de la

recherche medicale »

(INSERM)

1979–1985 Mortality RR (age, social

class, area of residence)

Men and women of all ages;

records of deaths in resident

population of France from 1979–

1985 (provided by INSERM)

Computation of differences in risk

between migrants using a case-

control approach

Country of birth Metropolitan-born

population in France

Migrants in population

data: 3.2%; Cancer

deaths among migrants:

n = 8708

Consumption of salted

and preserved foods

(nasopharyngeal cancer);

genetic susceptibility;

high and early exposure

to infection with

hepatitis-B virus and

aflatoxin, chronic

infection with liver flukes

(liver cancer)

Poor quality of French

mortality data; Small

number of deaths in

Chinese-born migrants

continued on next page

EU

RO

PE

AN

JO

UR

NA

LO

FC

AN

CE

R4

6(2

01

0)

26

47

–2

65

92

64

9

Ta

ble

1–

con

tin

ued

Co

un

try,

au

tho

rsa

nd

yea

r

of

stu

dy

Stu

dy

aim

:to

ex

plo

re

Da

taso

urc

ePeri

od

Ou

tco

me/m

ea

sure

of

ass

oci

ati

on

(cov

ari

able

s)

Co

ho

rta

cqu

isit

ion

/In

-a

nd

ex

clu

sio

ncr

iteri

a

Meth

od

olo

gic

al

pecu

lia

riti

es

Defi

nit

ion

of

eth

nic

ity

Refe

ren

cep

op

ula

tio

nS

ize

an

dco

mp

osi

tio

n

of

stu

dy

po

pu

lati

on

Dis

cuss

ed

ex

pla

na

tio

ns

for

risk

dif

fere

nce

s

Stu

dy

lim

ita

tio

ns

Germ

an

yS

pa

llek

et

al.

(2009)1

6

Ca

nce

rin

cid

en

cein

Turk

ish

imm

igra

nts

inH

am

bu

rg

Stu

dy

coh

ort

:H

am

bu

rg

Ca

nce

rR

egis

try

;

Refe

ren

cep

op

ula

tio

n:

po

pu

lati

on

regis

try

1990–2

004

Inci

den

ceR

R(y

ea

ro

f

bir

th)

Men

an

dw

om

en

of

all

ages;

iden

tifi

cati

on

of

case

so

fTu

rkis

h

ori

gin

by

use

of

na

me-b

ase

d

alg

ori

thm

Str

ati

fica

tio

nb

yb

irth

coh

ort

sto

ad

just

for

age

an

dto

inve

stig

ate

the

life

cou

rse

pers

pect

ive

Na

me-b

ase

d

alg

ori

thm

Rep

rese

nta

tive

po

pu

lati

on

sam

ple

of

Ha

mb

urg

Ca

nce

rca

ses

in

Turk

ish

mig

ran

ts:

n=

1346

Dif

fere

nt

nu

trit

ion

al

pa

ttern

s

(ca

nce

ro

fd

igest

ive,

uri

na

rytr

act

an

dp

rost

ate

);ea

rly

life

ex

peri

en

ces

an

din

fect

ion

s(e

.g.

HP

Va

nd

EB

V);

hig

her

smo

kin

gp

rev

ale

nce

am

on

g

Turk

ish

ma

les;

dif

fere

nt

rep

rod

uct

ive

beh

av

iou

r

Mis

cla

ssifi

cati

on

an

d

inco

mp

lete

iden

tifi

cati

on

can

no

tb

e

rule

do

ut

du

eto

na

me-

ba

sed

ap

pro

ach

;sm

all

nu

mb

er

of

Turk

ish

case

s;p

oss

ible

un

dere

stim

ati

on

du

eto

rem

igra

tio

n

Win

kle

ret

al.

(2009)3

5

Ca

nce

rm

ort

ali

ty

an

din

cid

en

cein

FS

Um

igra

nts

in

Germ

an

y

Ca

nce

rm

ort

ali

ty:

sam

ple

of

mig

ran

t

coh

ort

inN

ort

hR

hin

e-

West

ph

ali

a;

can

cer

inci

den

ce:

sam

ple

of

mig

ran

tco

ho

rtin

Sa

arl

an

d;

Lin

ka

ge

wit

h

loca

lp

op

ula

tio

n

regis

trie

s,lo

cal

rece

pti

on

cen

tres,

Sa

arl

an

dC

an

cer

Regis

try,

NR

Wst

ati

stic

al

offi

cea

nd

loca

lh

ea

lth

offi

ces

1990–2

005

SM

R,

SIR

(age,ca

len

da

r

yea

r)

Ca

nce

rm

ort

ali

ty:

arr

iva

lin

Germ

an

yb

etw

een

1st

Jan

ua

ry

1990

an

d31st

Dece

mb

er

2001;

aged

15+

;ca

nce

rin

cid

en

ce:

arr

iva

lb

etw

een

1990

an

d2005;

ex

clu

sio

n:

mis

sin

gd

ata

,ca

nce

r

dia

gn

osi

sin

cou

ntr

yo

fo

rigin

Da

teo

fa

rriv

al

ap

pro

xim

ate

db

y

pa

ssp

ort

issu

ed

ate

;p

ers

on

-yea

r

est

ima

tio

nu

sin

gG

erm

an

mo

rta

lity

rate

s

Rese

ttle

rsfr

om

FS

U

of

Germ

an

eth

nic

ity

Th

een

tire

po

pu

lati

on

of

Germ

an

y

Mo

rta

lity

:S

tud

y

(mig

ran

t)co

ho

rt:

n=

34,3

93,

Dea

ths

in

coh

ort

:n

=2580,

Ca

nce

rd

ea

ths:

n=

708;

Inci

den

ce:

Stu

dy

(mig

ran

t)co

ho

rt:

n=

18,6

19,

Ca

nce

r

case

s:n

=586

Sim

ila

rS

IRa

nd

SM

Rp

att

ern

s

(pre

sum

ab

lyn

osu

rviv

al

dif

fere

nce

s);

big

imp

act

of

smo

kin

g

pre

va

len

ce(s

imil

ar

toco

un

try

of

ori

gin

);H

.pyl

ori

pre

va

len

ce

(sto

ma

chca

nce

r);

alc

oh

ol

con

sum

pti

on

an

dh

ep

ati

tis

infe

ctio

n

(liv

er

can

cer)

;h

igh

er

bir

thra

tes

(bre

ast

can

cer)

Sli

gh

tly

dif

fere

nt

po

pu

lati

on

su

sed

for

sta

nd

ard

isa

tio

n;

foll

ow

-

up

est

ima

tio

n

Ro

nell

en

fits

ch

et

al.

(2009)3

6

Sto

ma

chca

nce

r

mo

rta

lity

inFS

U

mig

ran

tsin

Germ

an

y

Stu

dy

po

pu

lati

on

:

sam

ple

of

mig

ran

tsfr

om

FS

Uto

Germ

an

fed

era

l

sta

teN

ort

hR

hin

e

West

ph

ali

a,

iden

tifi

ed

in

mu

nic

ipa

lp

op

ula

tio

n

regis

trie

s;li

nk

age

wit

h

mo

rta

lity

da

ta(c

au

seo

f

dea

thd

ata

ba

se)

thro

ugh

sex

,d

ate

so

fb

irth

an

d

dea

th,

last

resi

den

cea

s

iden

tifi

ers

(reg

istr

y-

ba

sed

)

1990–2

005

SM

R(a

ge,

cale

nd

ar

yea

r)

Arr

iva

lin

Germ

an

yb

etw

een

1st

Jan

ua

ry1990

an

d31st

Dece

mb

er

2001;

aged

15+

;su

ccess

full

y

iden

tifi

ed

inele

ctro

nic

mu

nic

ipa

l

po

pu

lati

on

regis

trie

s

Foll

ow

-up

ass

ura

nce

thro

ugh

ele

ctro

nic

reco

rdli

nk

age

wit

h

mu

nic

ipa

lp

op

ula

tio

nre

gis

trie

s

an

da

sta

teca

use

of

dea

th

da

tab

ase

;v

ita

lst

atu

s

asc

ert

ain

men

t;ca

use

of

dea

th

retr

ieva

l

Rese

ttle

rsfr

om

FS

U

of

Germ

an

eth

nic

ity

Th

een

tire

po

pu

lati

on

of

Germ

an

y

Stu

dy

(mig

ran

t)co

ho

rt:

n=

34,3

93,

dea

ths

in

coh

ort

:n

=2580;

sto

ma

chca

nce

r

dea

ths:

n=

68

Lo

ng

late

ncy

aft

er

ex

po

sure

tori

sk

fact

ors

inea

rly

life

(e.g

.HP

infe

ctio

n;

con

tin

ua

tio

no

fli

fest

yle

sa

nd

beh

av

iou

rs(e

.g.

die

tary

ha

bit

s);

cha

nge

inh

ygie

ne

con

dit

ion

s,

ea

rlie

rd

ete

ctio

n,

bett

er

trea

tmen

t

op

tio

ns,

imp

rov

ed

acc

ess

to

hea

lth

care

Rest

rict

ed

da

ta

av

ail

ab

ilit

y;

dif

fere

nce

s

inst

ud

yp

op

ula

tio

ns;

no

info

rma

tio

no

nex

act

tum

ou

rlo

cati

on

Ott

et

al.

(2008)3

7

Mo

rta

lity

of

can

cers

of

po

ssib

ly

infe

ctio

us

ori

gin

in

mig

ran

tsfr

om

FS

U

toG

erm

an

y

Stu

dy

po

pu

lati

on

:

sam

ple

of

mig

ran

tsfr

om

FS

Uto

Germ

an

fed

era

l

sta

teN

ort

hR

hin

e

West

ph

ali

a,

iden

tifi

ed

in

mu

nic

ipa

lp

op

ula

tio

n

regis

trie

s;li

nk

age

wit

h

mo

rta

lity

da

ta(c

au

seo

f

dea

thd

ata

ba

se)

thro

ugh

sex

,d

ate

so

fb

irth

an

d

dea

th,

last

resi

den

cea

s

iden

tifi

ers

(reg

istr

y-

ba

sed

)

1990–2

005

SM

R(s

ex

,5-y

ea

ra

ge

gro

up

,ca

len

da

ry

ea

r,

len

gth

of

stay,

imm

igra

tio

np

eri

od

),

mo

rta

lity

RR

Arr

iva

lin

Germ

an

yb

etw

een

1st

Jan

ua

ry1990

an

d31st

Dece

mb

er

2001;

aged

15+

;su

ccess

full

y

iden

tifi

ed

inele

ctro

nic

mu

nic

ipa

l

po

pu

lati

on

regis

trie

s

Age-,

sex

-,ca

use

-a

nd

cale

nd

ar

yea

rsp

eci

fic

mo

rta

lity

rate

so

fth

e

Germ

an

po

pu

lati

on

ob

tain

ed

usi

ng

WH

O’s

Mo

rta

lity

Da

tab

ase

;E

ffect

of

len

gth

of

resi

den

cea

na

lysi

s

Rese

ttle

rsfr

om

FS

U

of

Germ

an

eth

nic

ity

Th

een

tire

po

pu

lati

on

of

Germ

an

y

Stu

dy

coh

ort

:

n=

34,3

93,

dea

ths

in

coh

ort

:n

=2580

H.

pylo

ria

nd

hep

ati

tis

vir

us

infe

ctio

n,

nu

trit

ion

al

fact

ors

(lo

w

fru

it/v

egeta

ble

con

sum

pti

on

,h

igh

inta

ke

of

nit

rite

-co

nta

inin

gfo

od

s),

hig

ha

lco

ho

lco

nsu

mp

tio

n(g

ast

ric

an

dli

ver

can

cer)

;li

vin

gco

nd

itio

ns;

dif

fere

nce

sin

hea

lth

-seek

ing

beh

av

iou

r

Resu

lts

no

ta

dju

sted

for

pre

va

len

ceo

fri

sk

fact

ors

;o

nly

lim

ited

inte

rpre

tati

on

of

resu

lts

po

ssib

leo

win

gto

ab

sen

ceo

feth

nic

-

speci

fic

mo

rta

lity

da

tain

stu

die

su

sin

g

ad

min

istr

ati

ve

da

ta

Ky

ob

utu

ngi

et

al.

(2006)3

8

Dif

fere

nce

sin

can

cer

mo

rta

lity

betw

een

eth

nic

Germ

an

imm

igra

nts

an

dth

en

ati

ve

Germ

an

po

pu

lati

on

Stu

dy

po

pu

lati

on

:

sam

ple

of

mig

ran

tsfr

om

FS

Uto

Germ

an

fed

era

l

sta

teN

ort

hR

hin

e

West

ph

ali

a,

iden

tifi

ed

in

mu

nic

ipa

lp

op

ula

tio

n

regis

trie

s;li

nk

age

wit

h

mo

rta

lity

da

ta(c

au

seo

f

dea

thd

ata

ba

se)

thro

ugh

sex

,d

ate

so

fb

irth

an

d

dea

th,

last

resi

den

cea

s

iden

tifi

ers

(reg

istr

y

ba

sed

)

1990–2

001/2

002

SM

R(a

ge,

cale

nd

ar

yea

r,a

rriv

al

peri

od

),

Mo

rta

lity

RR

Arr

iva

lin

Germ

an

yb

etw

een

1st

Jan

ua

ry1990

an

d31st

Dece

mb

er

2001;

aged

15+

;su

ccess

full

y

iden

tifi

ed

inele

ctro

nic

mu

nic

ipa

l

po

pu

lati

on

regis

trie

s

Age-,

sex

-,ca

use

-a

nd

cale

nd

ar

yea

rsp

eci

fic

mo

rta

lity

rate

so

fth

e

Germ

an

po

pu

lati

on

ob

tain

ed

usi

ng

WH

O’s

Mo

rta

lity

Da

tab

ase

;

an

aly

sis

of

secu

lar

tren

ds

an

d

eff

ect

of

len

gth

of

resi

den

ce;

dir

ect

lyst

an

da

rdis

es

dea

thra

tes

calc

ula

ted

for

all

-ca

nce

rsa

nd

lun

g

can

cer

Rese

ttle

rsfr

om

FS

U

of

Germ

an

eth

nic

ity

Th

een

tire

po

pu

lati

on

of

Germ

an

y

Pers

on

-yea

rso

fFU

in

mig

ran

tst

ud

yco

ho

rt:

n=

247,1

43;

Ca

nce

r

dea

ths

inco

ho

rt:

n=

469

Dif

fere

nce

sin

risk

fact

or

dis

trib

uti

on

:sm

ok

ing,

alc

oh

ol

con

sum

pti

on

,d

iet,

ph

ysi

cal

act

ivit

y,

rep

rod

uct

ive

his

tory

,h

ea

lth

care

uti

lisa

tio

n,

gen

eti

cfa

cto

rs,

vir

al

infe

ctio

ns;

can

cer

mo

rta

lity

ma

inly

infl

uen

ced

by

pre

-mig

rati

on

risk

fact

ors

(co

un

try

of

ori

gin

eff

ect

)

Ass

ess

men

to

fcu

rren

t

or

pre

-mig

rati

on

ind

ivid

ua

lri

skp

rofi

les

of

mig

ran

tsim

po

ssib

le;

inco

mp

lete

FU

for

som

e

coh

ort

mem

bers

Zeeb

et

al.

(2002)2

1

Th

etr

an

siti

on

in

can

cer

mo

rta

lity

pa

ttern

sa

mo

ng

Turk

ish

mig

ran

ts

resi

din

gin

Germ

an

y

Mo

rta

lity

da

ta:

dea

th

regis

trati

on

reco

rds

(fo

rmer)

West

Germ

an

y;

inci

den

ced

ata

:S

aa

rla

nd

can

cer

regis

try

1970–1

998

(In

cid

en

ce)

an

d

1980–1

997

(Mo

rta

lity

)

AS

MR

,P

CIR

(age)

Men

an

dw

om

en

aged

0–6

4;

use

of

na

me-b

ase

da

pp

roa

ch(b

ase

d

on

Turk

ish

firs

ta

nd

surn

am

es)

as

pro

xy

for

eth

nic

ity

Tim

etr

en

ds

for

AS

MR

sa

na

lyse

din

thre

eeq

ua

lin

terv

als

;m

issi

ng

info

rma

tio

no

neth

nic

ity

in

inci

den

td

en

om

ina

tor

po

pu

lati

on

rem

ed

ied

by

calc

ula

tio

no

fP

CIR

s

(no

min

ato

ro

nly

),ex

pect

ed

pro

po

rtio

ns

ob

tain

ed

by

use

of

stra

tifi

ed

ran

do

msa

mp

leo

fth

e

en

tire

regis

try

Na

tio

na

lity

(mo

rta

lity

an

aly

sis)

;n

am

e-

ba

sed

alg

ori

thm

(in

cid

en

cea

na

lysi

s)

Na

tive

Germ

an

po

pu

lati

on

Ca

nce

rd

ea

ths

am

on

g

mig

ran

ts:

n=

6054;

inci

den

tca

nce

rca

ses

am

on

gm

igra

nts

:

n=

163

Po

ten

tia

lri

skfa

cto

rs:

un

fav

ou

rab

le

livin

gco

nd

itio

ns

inch

ild

ho

od

,h

igh

pre

va

len

ceo

fH

.py

lori

infe

ctio

ns

am

on

gTu

rks

(sto

ma

chca

nce

r);h

igh

die

tary

en

erg

yin

tak

e(b

rea

st

can

cer)

;sm

ok

ing

tren

ds;

hep

ati

tis

B

infe

ctio

n(l

iver

can

cer)

;h

ea

lth

y

mig

ran

teff

ect

;re

-mig

rati

on

of

ill

mig

ran

ts(s

alm

on

bia

s);

life

sty

le

cha

nges;

soci

o-c

ult

ura

lb

arr

iers

aff

ect

ing

up

tak

ea

nd

qu

ali

tyo

f

clin

ica

ltr

ea

tmen

t

Stu

dy

rest

rict

ed

to

pers

on

sb

elo

w65;

sma

ll

nu

mb

er

of

case

s

beca

use

of

yo

un

ga

ge

dis

trib

uti

on

of

Turk

ish

mig

ran

tsin

Germ

an

y;

no

gen

era

tio

n

ass

ign

men

tp

oss

ible

an

d

sub

ject

tob

ias

(e.g

.

inte

rcu

ltu

ral

ma

rria

ges)

2650 E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

Sp

ain

Reg

ido

ret

al.

(2008)3

9

Wh

eth

er

mo

rta

lity

inim

mig

ran

tsin

the

regio

no

fM

ad

rid

dif

fers

fro

m

mo

rta

lity

inS

pa

nis

h

in-c

ou

ntr

ym

igra

nts

Mo

rta

lity

da

ta:

Mo

rta

lity

Reg

iste

r;p

op

ula

tio

n

da

ta:

Mu

nic

ipa

l

Po

pu

lati

on

Regis

ter,

cen

sus

da

ta(b

oth

sou

rces

pro

vid

ed

by

Ma

dri

dIn

stit

ute

of

Sta

tist

ics)

;u

nli

nk

ed

stu

dy

2000–2

004

Mo

rta

lity

RR

(age,

per

cap

ita

inco

me,

are

ao

f

resi

den

ce)

Men

aged

20–6

4Pe

rca

pit

ain

com

eest

ima

ted

ba

sed

on

inco

me

tax

retu

rns

for

the

yea

r

2000,

qu

art

iles

of

dis

trib

uti

on

ass

ign

ed

toea

chin

div

idu

al

ba

sed

on

cen

sus

tra

cto

fre

sid

en

ce

Co

un

try

of

bir

thS

pa

nis

hin

-co

un

try

mig

ran

ts;

po

pu

lati

on

bo

rnin

Ma

dri

d

Ca

nce

rd

ea

ths

am

on

g

mig

ran

ts:

n=

335

Hea

lth

y-m

igra

nt

eff

ect

;d

iffe

ren

ces

ind

em

ogra

ph

ics;

sta

ge

of

smo

kin

g

ep

idem

ic

Hete

rogen

eit

yw

ith

in

mig

ran

tgro

up

s;

info

rma

tio

no

n

po

pu

lati

on

an

dd

ea

ths

fro

md

iffe

ren

tso

urc

es

(nu

mera

tor/

den

om

ina

tor

info

rma

tio

nb

ias)

;n

o

info

rma

tio

no

nd

ura

tio

n

of

resi

den

ce

Sw

ed

en

Hem

min

ki

et

al.

(2010)4

0

Liv

er

an

d

ga

llb

lad

der

can

cer

inim

mig

ran

tsto

Sw

ed

en

Stu

dy

coh

ort

:S

wed

ish

Fam

ily

Ca

nce

rD

ata

ba

se;

lin

ka

ge

of

ad

min

istr

ati

ve

fam

ily

regis

ter

at

Sta

tist

ics

Sw

ed

en

toT

he

Sw

ed

ish

Ca

nce

rR

egis

try

1958–2

006

SIR

(5-y

ear

age

gro

up

,

sex

,p

eri

od

)

Fore

ign

-bo

rnm

en

an

dw

om

en

of

all

ages;

pri

ma

ryli

ver

can

cer

Co

un

try

of

bir

thN

ati

ve

Sw

ed

ish

po

pu

lati

on

Ca

nce

rca

ses

in

mig

ran

ts:

n=

1428

Ch

ron

icH

BV

infe

ctio

n,

oft

en

tra

nsm

itte

da

tb

irth

;li

ver

flu

ke

infe

ctio

ns;

po

or

liv

ing

con

dit

ion

s;

un

av

ail

ab

ilit

yo

fm

ed

ica

lca

re

Mo

usa

vi

et

al.

(2010)4

1

Ca

nce

rin

cid

en

cein

Ira

nia

nim

mig

ran

ts

toS

wed

en

Stu

dy

coh

ort

:S

wed

ish

Fam

ily

Ca

nce

rD

ata

ba

se

(cre

ate

db

yli

nk

age

of

ad

min

istr

ati

ve

fam

ily

regis

ter

at

Sta

tist

ics

Sw

ed

en

toT

he

Sw

ed

ish

Ca

nce

rR

egis

try

1958–2

006

SIR

(5-y

ear

age

gro

up

,

sex

,re

gio

n,

tim

e

peri

od

)

Men

an

dw

om

en

of

all

ages

Co

un

try

of

bir

thN

ati

ve

Sw

ed

ish

po

pu

lati

on

Ca

nce

rca

ses

in

mig

ran

ts:

n=

1293

En

vir

on

men

tal,

rep

rod

uct

ive

an

d

soci

oeco

no

mic

fact

ors

;H

ep

ati

tis

infe

ctio

nin

cou

ntr

yo

fo

rigin

;

smo

kin

g(b

lad

der

can

cer)

Mo

usa

vi

et

al.

(2010)4

2

Na

sop

ha

ryn

gea

l

an

d

hyp

op

ha

ryn

gea

l

can

cer

risk

in

imm

igra

nts

to

Sw

ed

en

Stu

dy

coh

ort

:S

wed

ish

Fam

ily

Ca

nce

rD

ata

ba

se

(cre

ate

db

yli

nk

age

of

ad

min

istr

ati

ve

fam

ily

regis

ter

at

Sta

tist

ics

Sw

ed

en

toT

he

Sw

ed

ish

Ca

nce

rR

egis

try

1958–2

006

SIR

(5-y

ear

age

gro

up

,

sex

,ti

me

peri

od

)

Men

an

dw

om

en

of

all

ages

Co

un

try

of

bir

thN

ati

ve

Sw

ed

ish

po

pu

lati

on

Ca

nce

rca

ses

in

mig

ran

ts:

n=

243

EB

Vin

fect

ion

inea

rly

life

;

dif

fere

nce

sin

smo

kin

ga

nd

die

tary

pa

ttern

s;ch

ewin

gto

ba

cco

Aze

rka

net

al.

(2008)4

3

Ris

ko

fin

va

siv

e

cerv

ica

lca

nce

r

am

on

gim

mig

ran

t

wo

men

Stu

dy

po

pu

lati

on

:

Sw

ed

ish

To

tal

Po

pu

lati

on

Regis

ter;

lin

ka

ge

wit

hS

wed

ish

Ca

nce

r,ca

use

of

dea

th

an

dm

igra

tio

nre

gis

ters

thro

ugh

na

tio

na

l

regis

tra

tio

nn

um

bers

1968–2

004

Inci

den

ceA

SR

,R

R(a

ge,

cale

nd

ar

peri

od

,S

ES

)

Wo

men

aged

13–7

9;

ex

clu

sio

n

crit

eri

a:

dea

th,

em

igra

tio

n,

his

tory

of

cerv

ica

lca

nce

rb

efo

re

en

try

into

coh

ort

;m

issi

ng

pla

ce

of

bir

tho

rm

igra

tio

nd

ate

;

wo

men

old

er

tha

n100

yea

rs

du

rin

gFU

Ca

tego

risa

tio

no

fm

igra

nt

ori

gin

s

into

low

-,m

ed

ium

-a

nd

hig

h-r

isk

are

as

for

cerv

ica

lca

nce

r,

acc

ou

nti

ng

for

inte

r-co

un

try

va

ria

tio

ns;

SE

So

bta

ined

fro

m1960,

1970,

1980

an

d1990

cen

suse

s,

cate

go

rise

din

tofi

ve

gro

up

s;eff

ect

of

du

rati

on

of

stay

(mo

reo

rle

ss

tha

n10

yea

rs);

stra

tifi

cati

on

by

age

at

imm

igra

tio

n

Bir

thre

gio

ns

Sw

ed

ish

-bo

rn

wo

men

Cerv

ica

lca

nce

rca

ses:

n=

19,5

42,

Ca

ses

am

on

gm

igra

nts

:

n=

1991

(10.2

%)

Ch

an

ges

inli

fest

yle

,se

xu

al

beh

av

iou

r;est

ab

lish

men

to

fce

rvic

al

can

cer

scre

en

ing

pro

gra

mm

es;

hea

lth

ym

igra

nt

eff

ect

;p

ers

iste

nt

HP

Vin

fect

ion

or

pre

can

cero

us

lesi

on

sb

efo

reim

mig

rati

on

Re-m

igra

tio

nw

ith

ou

t

reco

rdin

g,

lea

din

gto

un

dere

stim

ati

on

of

risk

s;y

ou

ng

mig

ran

t

po

pu

lati

on

s;h

igh

er

pro

po

rtio

no

f

un

cla

ssifi

ed

SE

Sa

mo

ng

imm

igra

nts

Mo

rad

iet

al.

(2008)4

4

Th

eo

ccu

rren

ceo

f

thyro

idca

nce

r

am

on

gS

wed

ish

resi

den

tsb

orn

in

Ira

nco

mp

are

dw

ith

tha

to

fS

wed

ish

-

bo

rnre

sid

en

ts

Stu

dy

coh

ort

:to

tal

Po

pu

lati

on

Regis

ter

held

by

Sta

tist

ics

Sw

ed

en

;

refe

ren

ceco

ho

rt:

Na

tio

na

lce

nsu

ses

1960–

1990,

lon

git

ud

ina

l

inte

gra

tio

nd

ata

ba

sefo

r

hea

lth

insu

ran

cea

nd

lab

ou

rm

ark

et

stu

die

s

1990–2

003;

lin

ka

ge

wit

h

Ca

nce

rR

egis

ter

thro

ugh

na

tio

na

lre

gis

trati

on

nu

mb

er

1969–2

004

Inci

den

ceR

R(a

ge,

cale

nd

ar

yea

r,

ed

uca

tio

n)

Men

an

dw

om

en

of

all

ages;

kn

ow

nd

ate

of

imm

igra

tio

n,

free

of

thy

roid

can

cer

at

sta

rto

fFU

Da

tao

np

are

nta

lp

lace

of

bir

th

thro

ugh

lin

ka

ge

wit

h

mu

ltig

en

era

tio

nre

gis

ter;

stra

tifi

cati

on

of

resu

lts

by

age

at

imm

igra

tio

n,

du

rati

on

of

stay

an

d

cale

nd

ar

yea

ro

fm

igra

tio

n(b

efo

re

or

aft

er

1990)

Co

un

try

of

bir

thN

ati

ve

Sw

ed

ish

po

pu

lati

on

Inci

den

tca

nce

rca

ses:

n=

9826;

am

on

g

mig

ran

ts:

n=

50

(0.5

%)

Ex

po

sure

toen

vir

on

men

tal

risk

fact

ors

du

rin

gea

rly

life

;io

din

e

defi

cien

cy;

hy

perp

last

icle

sio

ns

of

the

thy

roid

gla

nd

No

info

rma

tio

no

n

pre

va

len

ceo

fri

sk

fact

ors

;n

oin

form

ati

on

on

his

tolo

gic

al

cla

ssifi

cati

on

Hem

min

ki

et

al.

(2002)4

5

Ca

nce

rri

sks

in

ad

ult

imm

igra

nts

to

Sw

ed

en

Stu

dy

coh

ort

:S

wed

ish

Fam

ily

Ca

nce

rD

ata

ba

se

(cre

ate

db

yli

nk

age

of

ad

min

istr

ati

ve

fam

ily

regis

ter

at

Sta

tist

ics

Sw

ed

en

toT

he

Sw

ed

ish

Ca

nce

rR

egis

try

);

mo

rta

lity

da

ta:

dea

th

no

tifi

cati

on

da

ta;

ad

dit

ion

al

po

pu

lati

on

da

ta:

na

tio

na

lce

nsu

ses

of

1960,

1970,

1980

an

d

1990;

lin

ka

ge

thro

ugh

un

iqu

ete

chn

ica

l

iden

tifi

cati

on

nu

mb

er

1961–1

998

SIR

(5-y

ear

age

gro

up

,

sex

,re

gio

n,

peri

od

,

tum

ou

rty

pe)

Ad

ult

men

an

dw

om

en

;h

av

ing

chil

dre

nb

orn

inS

wed

en

(mem

ber

of

Fam

ily

Da

tab

ase

)

Co

un

try

of

bir

thS

wed

ish

na

tives

Ca

nce

rca

ses:

n=

673,4

24,

can

cer

case

sa

mo

ng

imm

igra

nts

:n

=32,7

22

(4.9

%)

Ma

rita

lst

atu

s;y

ou

ng

age

dis

trib

uti

on

of

imm

igra

nts

;to

ba

cco

con

sum

pti

on

(lu

ng,

uri

na

ry

bla

dd

er)

;p

igm

en

tati

on

,b

eh

av

iou

ral

dif

fere

nce

s(m

ela

no

ma

);

rep

rod

uct

ive

his

tori

es

(bre

ast

);

dia

gn

ost

ica

ctiv

ity

;m

ed

ica

lse

rvic

es

Mu

ltip

leco

mp

ari

son

pro

ble

m

Hem

min

ki

an

dLi

(2002)1

2

Ca

nce

rri

sks

in

Sw

ed

en

-bo

rn

desc

en

da

nts

of

imm

igra

nts

fro

m

Eu

rop

ea

na

nd

No

rth

Am

eri

can

cou

ntr

ies

Stu

dy

coh

ort

:S

wed

ish

Fam

ily

Ca

nce

rD

ata

ba

se

(cre

ate

db

yli

nk

age

of

ad

min

istr

ati

ve

fam

ily

regis

ter

at

Sta

tist

ics

Sw

ed

en

toT

he

Sw

ed

ish

Ca

nce

rR

egis

try

thro

ugh

un

iqu

ete

chn

ica

l

iden

tifi

cati

on

nu

mb

er)

1961–1

998

SIR

(5-y

ear

age

gro

up

,

sex

,re

gio

n,

peri

od

,

tum

ou

rty

pe)

Men

an

dw

om

en

aged

0–6

6S

epa

rate

an

aly

sis

by

fath

er’

sb

irth

cou

ntr

y,m

oth

er’

sb

irth

cou

ntr

y,fo

r

com

pa

trio

tp

are

nts

Pa

ren

tal

cou

ntr

yo

f

bir

th

Off

spri

ng

of

Sw

ed

ish

na

tiv

es

Ca

nce

rca

ses

by

fath

er’

sb

irth

cou

ntr

y:

n=

3460,

can

cer

case

s

by

mo

ther’

sb

irth

cou

ntr

y:

n=

4473

Lo

ng-l

ast

ing

en

vir

on

men

tal

an

d

heri

tab

leeff

ect

s(e

.g.

skin

pig

men

tati

on

);im

mu

ne

resp

on

se

Sm

all

nu

mb

er

of

case

s;

mu

ltip

leco

mp

ari

son

s

con

tin

ued

onn

ext

page

E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9 2651

Ta

ble

1–

con

tin

ued

Co

un

try,

au

tho

rsa

nd

yea

ro

f

stu

dy

Stu

dy

aim

:to

ex

plo

re

Da

taso

urc

ePeri

od

Ou

tco

me/m

ea

sure

of

ass

oci

ati

on

(cov

ari

able

s)

Co

ho

rta

cqu

isit

ion

/In

-a

nd

ex

clu

sio

ncr

iteri

a

Meth

od

olo

gic

al

pecu

lia

riti

es

Defi

nit

ion

of

eth

nic

ity

Refe

ren

cep

op

ula

tio

nS

ize

an

dco

mp

osi

tio

n

of

stu

dy

po

pu

lati

on

Dis

cuss

ed

ex

pla

na

tio

ns

for

risk

dif

fere

nce

s

Stu

dy

lim

ita

tio

ns

Th

e

Neth

erl

an

ds

Vis

ser

an

dva

n

Leeu

wen

(2007)4

6

Ca

nce

rri

skin

firs

t

gen

era

tio

nm

igra

nts

Po

pu

lati

on

da

ta:

an

nu

al

po

pu

lati

on

da

tafr

om

Sta

tist

ics

Neth

erl

an

ds;

lin

ked

wit

hS

tud

yC

oh

ort

:

Am

sterd

am

Ca

nce

r

Regis

try

(co

veri

ng

the

pro

vin

ces

No

rth

Ho

lla

nd

an

d

Fle

vo

lan

d);

po

pu

lati

on

-

ba

sed

1995–2

004

SIR

(age,

gen

der)

Men

an

dw

om

en

of

all

ages;

pri

ma

ryin

vasi

ve

can

cers

;

ex

clu

sio

ncr

iteri

on

:u

nk

no

wn

cou

ntr

yo

fb

irth

Co

un

try

of

bir

th(i

fp

oss

ible

)

ver

ified

wit

hin

form

ati

on

fro

m

na

tio

na

lp

op

ula

tio

nn

etw

ork

(e.g

.

scre

en

ing

pa

rtic

ipa

nts

)

Resi

den

tsb

orn

ou

tsid

eth

e

Neth

erl

an

ds

Na

tive

Du

tch

po

pu

lati

on

of

No

rth

Ho

lla

nd

/Fle

vo

an

d

Ca

nce

rca

ses:

n=

106,4

15;

can

cer

case

sa

mo

ng

mig

ran

ts:

n=

9271

(9%

)

Ex

po

sure

toin

fect

iou

sd

isea

ses

befo

rem

igra

tio

n;

hea

lth

yli

fest

yle

hab

its

pro

tect

ing

aga

inst

can

cer;

gen

eti

cfa

cto

rs(e

.g.

hig

her

pro

sta

te

can

cer

risk

inS

uri

na

mese

ma

les)

Sele

ctiv

e(r

e-

)mig

rati

on

Sti

rbu

et

al.

2006

(14)

Dif

fere

nce

sin

can

cer

mo

rta

lity

betw

een

mig

ran

ts

an

dth

en

ati

ve

Du

tch

po

pu

lati

on

Po

pu

lati

on

da

ta:

mu

nic

ipa

lp

op

ula

tio

n

regis

ters

;li

nk

ed

thro

ugh

pers

on

al

iden

tifi

cati

on

nu

mb

ers

tom

ort

ali

tyd

ata

:

cau

seo

fd

ea

thre

gis

try

(po

pu

lati

on

-ba

sed

)

1995–2

000

Mo

rta

lity

RR

(age,

sex

,

ma

rita

lst

atu

s,

urb

an

isa

tio

nle

vel,

are

ain

com

e)

Men

an

dw

om

en

of

all

ages;

lega

l

resi

den

tso

fth

eN

eth

erl

an

ds

Age

at

imm

igra

tio

na

nd

du

rati

on

of

resi

den

ceb

ase

do

nla

test

kn

ow

n

da

teo

fim

mig

rati

on

;d

egre

eo

f

urb

an

isa

tio

na

nd

mea

nh

ou

seh

old

eq

uiv

ale

nt

use

dto

ap

pro

xim

ate

SE

Sca

lcu

late

db

ase

do

np

ost

al

cod

e

Resi

den

tso

rp

are

nts

of

resi

den

tsb

orn

ab

roa

d(p

red

om

ina

nt

role

of

cou

ntr

yo

f

bir

tho

fm

oth

er)

Na

tive

Du

tch

po

pu

lati

on

Ca

nce

rd

ea

ths:

n=

173,4

61,

dea

ths

am

on

gm

igra

nts

n=

1454

(0.8

%)

Hea

lth

y-m

igra

nt/

un

hea

lth

y-

rem

igra

nt

eff

ect

;u

pta

ke

of

west

ern

life

sty

le(s

mo

kin

g,

cha

nges

ind

iet

an

dre

pro

du

ctiv

eb

eh

av

iou

r);

hep

ati

tis

Bsu

rfa

cea

nti

gen

s(r

isk

fact

or

for

liver

can

cer)

;im

po

rta

nce

of

life

-co

urs

ep

ers

pect

ive

Lim

ited

sta

tist

ica

l

po

wer

ow

ing

tosm

all

nu

mb

ers

an

dre

lati

vely

yo

un

g(a

nd

hig

hly

dif

fere

nt)

age

dis

trib

uti

on

so

f

mig

ran

ts

Bo

set

al.

(2004)4

7

Fact

ors

cau

sin

ga

hig

her

or

low

er

mo

rta

lity

in

mig

ran

tsco

mp

are

d

wit

hth

en

ati

ve

po

pu

lati

on

Po

pu

lati

on

da

ta:

mu

nic

ipa

lp

op

ula

tio

n

regis

ters

;li

nk

ed

thro

ugh

pers

on

al

iden

tifi

cati

on

nu

mb

ers

tom

ort

ali

tyd

ata

:

cau

seo

fd

ea

thre

gis

try

(po

pu

lati

on

-ba

sed

)

1995–2

000

Mo

rta

lity

RR

(age,

ma

rita

lst

atu

s,re

gio

n,

degre

eo

fu

rba

nis

ati

on

,

SE

Sb

yse

x)

Men

an

dw

om

en

of

all

ages;

lega

l

resi

den

tso

fth

eN

eth

erl

an

ds

Co

un

try

of

bir

tho

f

sub

ject

an

db

oth

pa

ren

ts(n

on

-Du

tch

if

at

lea

sto

ne

pa

ren

t

bo

rna

bro

ad

)

Na

tive

Du

tch

po

pu

lati

on

All

dea

ths

inD

utc

h

po

pu

lati

on

du

rin

g

stu

dy

peri

od

:

n=

750,1

48

Hea

lth

y-m

igra

nt/

un

hea

lth

y-

rem

igra

nt

eff

ect

;sm

ok

ing,

die

tary

hab

its

(ad

ap

tati

on

of

un

hea

lth

y

wes

tern

life

sty

le)

No

info

rma

tio

no

n

wit

hin

-mig

ran

tgro

up

va

ria

tio

ns;

risk

un

dere

stim

ati

on

in

som

egro

up

s;

un

regis

tere

d

rem

igra

tio

n

Vis

ser

et

al.

(2004)4

8

Bre

ast

can

cer

inci

den

cein

mig

ran

tsin

the

Neth

erl

an

ds

Po

pu

lati

on

da

ta:

an

nu

al

po

pu

lati

on

da

tao

bta

ined

fro

m

Sta

tist

ics

Neth

erl

an

ds;

stu

dy

coh

ort

:

Am

sterd

am

Ca

nce

r

Regis

try

an

dC

an

cer

Cen

tre

West

(co

veri

ng

the

pro

vin

ces

No

rth

Ho

lla

nd

an

dT

he

Ha

gu

e)

lin

ked

to

scre

en

ing

da

ta

1989–1

998

SIR

Wo

men

of

all

ages

Va

lid

ati

on

of

cou

ntr

yo

fb

irth

info

rma

tio

nw

ith

bre

ast

can

cer

scre

en

ing

pro

gra

mm

es

toC

an

cer

regis

try

da

ta;

ifd

ata

inca

nce

r

regis

try

dis

crep

an

to

rm

issi

ng,

cou

ntr

yo

fb

irth

info

rma

tio

nfr

om

scre

en

ing

da

tau

sed

Co

un

try

of

bir

thN

ati

ve

Du

tch

wo

men

Ca

nce

rca

ses:

n=

20,0

16

(am

on

g

mig

ran

ts:

n=

1699

(8.5

%)

Scr

een

ing

att

en

da

nce

;ch

an

ge

in

rep

rod

uct

ive

risk

fact

ors

such

as

low

er

pa

rity

Vis

ser

et

al.

(2003)4

9

Inci

den

ceo

f

cerv

ica

lca

nce

rin

No

rth

Ho

lla

nd

by

cou

ntr

yo

fb

irth

Po

pu

lati

on

da

ta:

an

nu

al

po

pu

lati

on

da

tao

bta

ined

fro

m

Sta

tist

ics

Neth

erl

an

ds;

stu

dy

coh

ort

:

Am

sterd

am

Ca

nce

r

Regis

try

(co

veri

ng

the

pro

vin

ces

No

rth

Ho

lla

nd

an

dFle

vo

lan

d)

1988–1

998

AS

IR,

O/E

rati

o(a

ge)

Wo

men

of

all

ages

wit

hin

va

sive

cerv

ica

lca

nce

r

Du

tch

resi

den

tb

orn

ab

roa

d

Na

tive

Du

tch

wo

men

Ca

nce

rca

ses:

n=

1530

(am

on

gm

igra

nts

:

n=

232

(15.2

%)

HP

Vin

fect

ion

;ch

an

ges

inli

fest

yle

;

scre

enin

gp

rogra

mm

es

inh

ost

cou

ntr

y;

sele

ctio

neff

ect

s

Mis

sin

gco

un

try

of

bir

thin

10%

of

case

s;

inco

mp

lete

ness

of

mo

rta

lity

regis

trati

on

;

no

info

rma

tio

no

n

pre

va

len

ceo

fri

sk

fact

ors

an

dd

iffe

ren

ces

inS

ES

Un

ited

Kin

gd

om

Ha

rdin

get

al.

(2009)2

8

Tre

nd

sin

can

cer

mo

rta

lity

in

mig

ran

tsli

vin

gin

En

gla

nd

an

dW

ale

s

An

on

ym

ised

dea

th

reco

rds;

po

pu

lati

on

da

tafr

om

the

1981,

1991

an

d2001

cen

suse

sfo

rE

ngla

nd

an

dW

ale

s

1979–2

003

Mo

rta

lity

RR

(age)

Men

an

dw

om

en

aged

30–6

9;

con

sist

en

tco

un

try

of

bir

th

defi

nit

ion

inb

oth

dea

ths

an

d

cen

sus

da

ta

Tre

nd

an

aly

sis

(ch

an

ges

ind

ea

th

rate

sa

mo

ng

thre

eti

me

inte

rva

ls)

Co

un

try

of

bir

thE

ngli

sh-a

nd

Wels

h-

bo

rn

Ch

an

ges

inri

skb

eh

av

iou

r

(co

nv

erg

en

cein

rate

sto

tho

seo

f

En

gla

nd

an

dW

ale

s);

risi

ng

smo

kin

g

tren

ds

am

on

gim

mig

ran

ts;

alc

oh

ol

con

sum

pti

on

;d

ela

yed

up

tak

ea

nd

po

ore

rq

ua

lity

of

clin

ica

l

ma

na

gem

en

t;p

oo

rca

nce

r

aw

are

ness

;co

-mo

rbid

itie

s;h

isto

ric

(vir

al)

infe

ctio

ns

Po

ssib

le

mis

cla

ssifi

cati

on

of

cou

ntr

yo

fb

irth

betw

een

cen

sus

da

ta

an

dd

ea

thce

rtifi

cate

s;

hea

lth

y-m

igra

nt

eff

ect

(sele

ctio

nb

ias)

Jack

et

al.

(2009)1

7

Bre

ast

can

cer

inci

den

ce,

sta

ge,

trea

tmen

ta

nd

surv

iva

lin

eth

nic

gro

up

sin

So

uth

Ea

st

En

gla

nd

Stu

dy

coh

ort

:T

ha

mes

Ca

nce

rR

egis

try

;

Na

tio

na

lH

ea

lth

Serv

ice

Cen

tra

l

Regis

ter;

po

pu

lati

on

da

ta:

Offi

cefo

r

Na

tio

na

lS

tati

stic

s

(ma

tch

ing

on

NH

S

nu

mb

er)

;re

gis

try

-a

nd

po

pu

lati

on

-ba

sed

stu

dy

1998–2

003

Inci

den

ceR

R,

HR

(age,

soci

oeco

no

mic

dep

riv

ati

on

,st

age

at

dia

gn

osi

s,tr

ea

tmen

t)

Wo

men

of

all

ages;

kn

ow

n

eth

nic

ity,

com

ple

tere

gis

trati

on

info

rma

tio

n;

ex

clu

sio

ncr

iteri

a:

pa

tien

tsre

gis

tere

db

yd

ea

th

cert

ifica

teo

nly

ex

clu

ded

fro

m

an

aly

ses

on

sta

ge,

trea

tmen

ta

nd

mo

rta

lity

So

cio

-dem

ogra

ph

icd

ep

riva

tio

n

ba

sed

on

inco

me

do

ma

ino

fIn

dex

of

Mu

ltip

leD

ep

riv

ati

on

2000,

div

ided

into

qu

inti

les,

ass

ign

ed

to

reco

rds

usi

ng

po

stco

de

of

resi

den

cea

td

iagn

osi

s

Self

-ass

ign

ed

eth

nic

ity

(usi

ng

cod

es

fro

m1991

an

d2001

cen

suse

s)

Wh

ite

wo

men

Ca

nce

rco

ho

rt:

n=

33,0

24

Scr

een

ing

up

tak

e;

trea

tmen

t

dif

fere

nce

s;re

pro

du

ctiv

e,

soci

oeco

no

mic

,a

nth

rop

om

etr

ica

nd

die

tary

fact

ors

;d

iffe

ren

ces

in

dis

ea

sep

erc

ep

tio

na

nd

resu

ltin

g

acc

ess

toh

ea

lth

care

serv

ices

Eth

nic

ity

info

rma

tio

n

no

tav

ail

ab

lefo

rla

rge

po

rtio

no

fp

ati

en

ts

(36%

);

rep

rese

nta

tiv

en

ess

of

eth

nic

gro

up

s;w

ith

in

eth

nic

gro

up

va

ria

tio

n

2652 E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

Wil

det

al.

(2006)2

2

Ca

nce

rm

ort

ali

tyin

En

gla

nd

an

dW

ale

s

by

cou

ntr

yo

fb

irth

Pop

ula

tio

nd

ata

:

Na

tio

na

lS

tati

stic

s,2001

Cen

sus

of

En

gla

nd

an

d

Wa

les;

mo

rta

lity

da

ta:

Offi

ceo

fN

ati

on

al

Sta

tist

ics

2001–2

003

SM

R(a

ge)

Men

an

dw

om

en

aged

20+

Co

un

try

of

bir

thE

ngla

nd

an

dW

ale

sa

s

aw

ho

le

Ca

nce

rd

ea

ths:

n=

398,5

15;

am

on

g

no

n-w

este

rnm

igra

nts

:

n=

13,1

61

(3.3

%)

Co

mp

lex

com

bin

ati

on

of

gen

eti

c

an

den

vir

on

men

tal

fact

ors

(die

t,

life

sty

le,

soci

oeco

no

mic

sta

tus)

;

smo

kin

gh

ab

its

(lu

ng

can

cer)

;p

oo

r

up

tak

eo

fca

nce

rsc

reen

ings;

un

ex

pla

ined

hig

hri

sks

inW

est

Afr

ica

ns

No

info

rma

tio

no

n

en

vir

on

men

tal

an

d

dem

ogra

ph

icfa

cto

rs;

lim

ited

reli

ab

ilit

yo

f

cou

ntr

yo

fb

irth

as

eth

nic

ity

pro

xy

;

nu

mera

tor/

den

om

ina

tor

bia

s;a

ccu

racy

of

cau

se

of

dea

thin

form

ati

on

;

po

ssib

le

mis

cla

ssifi

cati

on

of

cou

ntr

yo

fb

irth

Sm

ith

et

al.

(2003)1

8

Rece

nt

tren

ds

in

can

cer

inci

den

ce

am

on

gU

KS

ou

th

Asi

an

s

Pop

ula

tio

nd

ata

:

est

ima

tes

fro

m1991

cen

sus

of

En

gla

nd

an

d

Wa

les;

Stu

dy

coh

ort

:

Tre

nt

Ca

nce

rR

egis

try

1990–1

999

Inci

den

ceR

R(a

ge,

dep

riva

tio

n)

Men

an

dw

om

en

of

all

ages

(fo

r

all

-ca

nce

r)/a

ged

30+

(fo

rsi

te-

speci

fic

an

aly

ses)

Leve

lo

fd

ep

riv

ati

on

of

the

pa

tien

t’s

are

ao

fre

sid

en

ce(S

ES

pro

xy

)u

sin

g

To

wn

sen

dIn

dex

Su

r-a

nd

fore

na

me

En

gli

shn

on

-So

uth

Asi

an

s

Ca

nce

rca

ses:

n=

12,1

28;

am

on

g

mig

ran

ts(n

=862

(7.1

%)

Ha

rdin

ga

nd

Ro

sato

(1999)5

0

Inci

den

ceo

fca

nce

rs

am

on

gfo

reig

n-b

orn

resi

den

tso

fE

ngla

nd

an

dW

ale

s

Stu

dy

coh

ort

:1%

sam

ple

of

the

po

pu

lati

on

of

En

gla

nd

an

dW

ale

s;

lin

ked

toC

an

cer

regis

trati

on

s:N

HS

Cen

tra

lR

egis

ter

1971–1

989

SIR

(sex

,a

ge,

yea

ro

f

dia

gn

osi

s)

Men

an

dw

om

en

aged

15+

Cla

ssifi

cati

on

of

resu

lts

by

reli

gio

nC

ou

ntr

yo

fb

irth

,

en

ha

nce

db

yn

am

e

an

aly

sis,

cla

ssifi

ed

by

reli

gio

n

All

mem

bers

of

the

stu

dy

Ca

nce

rca

ses

am

on

g

no

n-w

este

rnm

igra

nts

:

n=

167

Dif

fere

nce

sin

soci

oeco

no

mic

sta

tus;

smo

kin

ga

nd

alc

oh

ol

con

sum

pti

on

;d

ieta

ryh

ab

its;

up

tak

e

of

scre

enin

gse

rvic

es;

life

sty

le;

occ

up

ati

on

Un

kn

ow

nem

igra

tio

ns;

hig

her

loss

of

FU

am

on

g

mig

ran

ts;

self

-sele

ctio

n

pro

cess

es;

no

info

rma

tio

no

nso

cia

l

dis

trib

uti

on

of

risk

fact

ors

Haw

ort

h

et

al.

(1999)5

1

Mo

rta

lity

fro

m

cirr

ho

sis

of

the

liver

an

dp

rim

ary

liver

can

cer

infi

rst-

gen

era

tio

nm

igra

nts

toE

ngla

nd

an

d

Wa

les

Pop

ula

tio

nd

ata

:

est

ima

tes

fro

m1991

cen

sus

of

En

gla

nd

an

d

Wa

les;

mo

rta

lity

da

ta:

Offi

cefo

rN

ati

on

al

Sta

tist

ics

1988–1

992

SM

R(a

ge,

sex

)M

en

an

dw

om

en

aged

20–6

9C

ou

ntr

yo

fb

irth

Na

tive

po

pu

lati

on

of

En

gla

nd

an

dW

ale

s

Ca

nce

rd

ea

ths:

n=

3237;C

an

cer

dea

ths

am

on

gm

igra

nts

fro

m

no

n-w

este

rn

cou

ntr

ies:

n=

238

(7.4

%)

Cu

ltu

re;

Reli

gio

n;

So

cio

eco

no

mic

dif

fere

nce

s;A

lco

ho

lin

tak

e;

Ch

ron

ic

hep

ati

tis

Ba

nd

Cin

fect

ion

s;li

fest

yle

Sm

all

nu

mb

ers

of

dea

ths

(esp

eci

all

y

am

on

gfe

ma

les)

Win

ter

et

al.

(1999)2

0

Ca

nce

rin

cid

en

cein

the

So

uth

Asi

an

po

pu

lati

on

of

En

gla

nd

Pop

ula

tio

nd

ata

:

est

ima

tes

fro

m1991

cen

sus

of

En

gla

nd

an

d

Wa

les;

stu

dy

coh

ort

:

Ca

nce

rR

egis

trie

so

f

Th

am

es,

Tre

nt,

West

Mid

lan

ds

an

dY

ork

shir

e

1990–1

992

AS

IR(a

ge)

Men

an

dw

om

en

of

all

ages

So

uth

Asi

an

na

mes

iden

tifi

ed

usi

ng

aco

mp

ute

rp

rogra

mm

e;

two

com

pa

riso

ns:

So

uth

Asi

an

vers

us.

no

n-S

ou

thA

sia

n(E

ngla

nd

)a

nd

So

uth

Asi

an

sin

En

gla

nd

vers

us

da

tafr

om

the

Ind

ian

sub

con

tin

en

t

Eth

nic

ori

gin

dete

rmin

ed

ba

sed

on

na

mes

No

n-S

ou

thA

sia

n

En

gli

shp

op

ula

tio

no

f

stu

dy

regio

n;

Ind

ian

regis

try

da

ta

Inci

den

tca

nce

rca

ses:

n=

356,5

55;

Ca

ses

am

on

gm

igra

nts

:

n=

3845

(1.1

%)

Lif

est

yle

;d

iet;

acc

ess

an

du

pta

ke

of

hea

lth

serv

ices

(scr

een

ings)

;

chew

ing

of

tob

acc

ow

ith

bete

l-q

uid

(ris

kfa

cto

rfo

rca

nce

ro

fto

ngu

e,

mo

uth

an

dh

yp

op

ha

ryn

x);

hep

ati

tis

Bin

fect

ion

(liv

er)

Wit

hin

-gro

up

dif

fere

nce

sw

ith

rega

rd

toli

fest

yle

,d

iet

etc

.;

po

ssib

le

mis

cla

ssifi

cati

on

of

na

mes;

very

lim

ited

va

lid

ity

of

com

pa

riso

n

betw

een

En

gli

shS

ou

th

Asi

an

rate

sa

nd

Ind

ian

Asi

an

rate

s

Sw

erd

low

et

al.

(1995)1

9

Ca

nce

rri

sks

in

Bri

tish

eth

nic

an

d

Ind

ian

eth

nic

mig

ran

tsto

En

gla

nd

an

dW

ale

s

Mo

rta

lity

da

ta:

Offi

ceo

f

Pop

ula

tio

nC

en

suse

s

an

dS

urv

eys

1973–1

985

Mo

rta

lity

RR

(MH

OR

)

(age)

Men

an

dw

om

en

of

all

ages;

ex

clu

sio

ncr

iteri

on

:u

nk

no

wn

eth

nic

ity

No

suit

ab

led

en

om

ina

tor

info

rma

tio

n:

calc

ula

tio

no

fo

dd

s

rati

os,

risk

of

dea

thfr

om

ea

ch

can

cer

site

inea

chm

igra

nt

gro

up

/

risk

of

dea

thfr

om

the

sam

eca

nce

r

site

inE

ngli

sh-

an

dW

els

h-b

orn

resi

den

ts(r

ela

tive

risk

of

mo

rta

lity

);In

form

ati

on

on

soci

al

cla

ss,

ma

rita

lst

atu

sa

nd

pa

rity

fro

m1971

cen

sus

Co

un

try

of

bir

tha

nd

eth

nic

gro

up

(dete

rmin

ed

on

the

ba

sis

of

na

mes)

Na

tive

En

gli

sha

nd

Wels

hn

ati

ves

Ca

nce

rd

ea

ths

am

on

g

mig

ran

ts:

n=

8282

Dif

fere

nce

sin

ex

po

sure

s

(occ

up

ati

on

);d

iffe

ren

ces

inso

cia

l

cla

ss;

bete

l-q

uid

chew

ing

(ora

la

nd

ph

ary

ngea

lca

nce

rs);

smo

kin

ga

nd

alc

oh

ol

con

sum

pti

on

(oeso

ph

agea

l

an

dla

ryn

gea

lca

nce

r);

hep

ati

tis

B

infe

ctio

n(l

iver)

;o

besi

ty

(ga

llb

lad

der)

;re

pro

du

ctiv

efa

cto

rs

(bre

ast

)

Nu

mera

tor

on

ly

mea

sure

;re

lia

bil

ity

of

eth

nic

cod

ing;

mis

reco

rdin

go

fd

ea

th

cert

ifica

tes;

no

info

rma

tio

no

n

dis

trib

uti

on

of

risk

fact

ors

Gru

lich

et

al.

(1992)2

3

Sit

e-s

peci

fic

can

cer

mo

rta

lity

inW

est

Afr

ica

n-,

Ea

st

Afr

ica

n-

an

d

Ca

rib

bea

n-b

orn

imm

igra

nts

Pop

ula

tio

nd

ata

:1971

po

pu

lati

on

cen

sus

of

En

gla

nd

an

dW

ale

s;

mo

rta

lity

da

ta:

Offi

ceo

f

Pop

ula

tio

nC

en

suse

s

an

dS

urv

eys

1970–1

985

Mo

rta

lity

RR

(age,

cale

nd

ar

peri

od

,so

cia

l

cla

ss)

Men

an

dw

om

en

of

all

ages

So

cia

lcl

ass

reco

rded

on

dea

th

cert

ifica

tes

(on

lyav

ail

ab

lefo

r

sub

ject

saged

15–6

4),

on

ly

perf

orm

ed

for

Ca

rib

bea

n

Co

un

try

of

bir

thE

ngla

nd

an

dW

ale

s

na

tives

Ca

nce

rd

ea

ths

am

on

g

mig

ran

ts:

n=

5407

Vir

al

ori

gin

s;d

iffe

ren

ces

inso

cia

l

cla

ss;

west

ern

life

sty

le;

gen

eti

c

dis

po

siti

on

;h

ep

ati

tis

Bin

fect

ion

(liv

er)

;d

iagn

ost

icfa

cili

ties;

bete

l-

chew

ing

(ora

lca

nce

r)

Sw

erd

low

(1991)5

2

Ca

nce

rin

cid

en

ce

an

dm

ort

ali

tyin

Vie

tna

mese

refu

gees

inE

ngla

nd

an

dW

ale

s

Stu

dy

po

pu

lati

on

:

Na

tio

na

lH

ea

lth

Serv

ice

Cen

tra

lR

egis

ter

1979–1

989

SM

R,

SR

R(a

ge)

Ma

lea

nd

fem

ale

Vie

tna

mese

refu

gees

bo

rnb

efo

re1950

an

d

wit

hN

HS

regis

trati

on

sfr

om

24

May

1979

to16

July

1985;

ex

clu

sio

ncr

iteri

a:

un

kn

ow

nse

x,

dea

thd

uri

ng

FU

Ob

serv

ed

mo

rta

lity

an

dca

nce

r

regis

trati

on

sco

mp

are

dw

ith

age-,

sex

-,a

nd

yea

r-sp

eci

fic

ex

pect

ati

on

sd

eri

ved

fro

m

ap

pli

cati

on

of

En

gla

nd

an

dW

ale

s

na

tio

na

lm

ort

ali

tya

nd

can

cer

regis

trati

on

rate

sto

the

pers

on

-

yea

rsa

tri

sk

Refu

gees

bo

rna

bro

ad

Mo

rta

lity

an

dca

nce

r

inci

den

cein

En

gla

nd

an

dW

ale

s(c

an

cer

regis

tra

tio

nd

ata

)

Mig

ran

tst

ud

yco

ho

rt:

n=

3327;

tota

l

mo

rta

lity

:n

=187;

can

cer

inci

den

ce:

n=

49

Hig

hp

reva

len

ceo

ftu

berc

ulo

sis

an

d

hep

ati

tis

B

Inco

mp

lete

dea

th

regis

trati

on

;se

lect

ion

bia

s

Ba

rker

an

d

Ba

ker

(1990)1

5

Inci

den

ceo

fca

nce

r

inA

sia

ns

liv

ing

in

Bra

dfo

rd,

En

gla

nd

Pop

ula

tio

nd

ata

:1981

cen

sus

(Un

ivers

ity

of

Leed

s);

stu

dy

po

pu

lati

on

:Y

ork

shir

e

Reg

ion

al

Ca

nce

rR

egis

try

1979–1

984

Inci

den

ceS

RR

(age)

Men

an

dw

om

en

of

all

ages

Cen

sus

da

tare

stri

cted

toh

ea

ds

of

ho

use

ho

lds;

Ca

nce

rri

ska

mo

ng

Asi

an

sin

Bra

dfo

rdco

mp

are

dw

ith

no

n-A

sia

ns

inB

rad

ford

an

dw

ith

can

cer

regis

try

da

tafr

om

Bo

mb

ay

Eth

nic

ori

gin

dete

rmin

ed

on

the

ba

sis

of

na

me

Ra

tes

of

no

n-A

sia

n

po

pu

lati

on

/th

e

Bo

mb

ay

can

cer

regis

trie

s

Ca

nce

rca

ses

am

on

g

mig

ran

ts:

n=

178

Ch

an

ge

inen

vir

on

men

tal

an

d

beh

av

iou

ral

infl

uen

cesc

ing,

hea

vy

alc

oh

ol

con

sum

pti

on

,b

ete

l

chew

ing,

con

sum

pti

on

of

spic

ed

foo

ds

(hy

po

ph

ary

nx

,p

ha

ryn

x);

up

tak

eo

fca

nce

rsc

reen

ings

Sm

all

nu

mb

er

of

case

s;

lim

ited

va

lid

ity

of

com

pa

riso

nw

ith

Bo

mb

ayra

tes;

mis

cla

ssifi

cati

on

of

eth

nic

ity

AS

MR

,age-s

tan

da

rdis

ed

mo

rta

lity

rate

/ra

tio

;AS

IR,a

ge-s

tan

da

rdis

ed

inci

den

cera

te/r

ati

o;A

SR

,age-s

tan

da

rdis

ed

rate

;E

BV

,Ep

stein

–Ba

rrv

iru

s;FS

U,F

orm

er

So

vie

tU

nio

n;F

U,f

oll

ow

-up

;HP

V,H

um

an

Pap

illo

ma

Vir

us;

NH

L,n

on

-Ho

dgk

in’s

lym

ph

om

a;N

HS

,Na

tio

na

lH

ea

lth

Serv

ice;O

/E-r

ati

o,o

bse

rved

/ex

pect

ed

rati

o;O

R,o

dd

sra

tio

;PC

IR,p

rop

ort

ion

al

can

cer

inci

den

cera

tio

;R

R,

risk

/ra

tera

tio

;S

ES

,so

cio

eco

no

mic

sta

tus;

SIR

,st

an

da

rdis

ed

inci

den

cera

te/r

ati

o;

SM

R,

sta

nd

ard

ised

mo

rta

lity

rate

/ra

tio

;S

RR

,st

an

da

rdis

ed

regis

tra

tio

nra

te/r

ati

o.

E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9 2653

2654 E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

Only one study focused entirely on second-generation mi-

grants12 (based on the patient’s own and parental country of

birth) and two other studies included this group explicitly in

addition to first-generation migrants.13,14 Seven studies in-

cluded descendants indirectly, owing to the method used

for identifying migrants.15–21 For instance, the name-based

approach does not allow a distinction between generations,

which can only be estimated vaguely based on age. There

were 27 studies that were aimed at first-generation migrants

only.

For reasons of clarity, migrant origins have been pooled

into the following categories: Eastern Europe [Former Soviet

Union (FSU), Russia], Africa (North, West and East Africa), Mid-

dle East (most frequently referring to Iran, Iraq and adjacent

countries), Southern Europe/Turkey, Asia [divided into general

Asia (mostly China and Vietnam) and South Asia (including In-

dia, Bangladesh, Indonesia, Ceylon and Pakistan)] and South-

ern and Central America. Owing to inconsistent definitions

between the studies, some overlap cannot be excluded.

2.4. Applied methods

Studies investigating cancer incidence used mainly cancer

registry data (21/37). Studies assessing cancer mortality drew

mostly on vital statistics such as mortality or cause of death

registries and databases or surveys (17/37). Population data

were obtained from population registers/statistical bureaux

(17/37), census data (13/37) – which were primarily used in

studies from France and the United Kingdom (UK) – or a

population sample (7/37). Most studies were population-/

registry-based. In many studies linkage procedures had been

performed using a unique identifier such as the ‘Personal

Identity Number’ in Sweden and the ‘National Health Service

(NHS) number’ in the UK. Two studies used numerator-only

analyses.

Some studies adjusted for a socioeconomic proxy and also

took important covariables such as duration of stay, age at

immigration and calendar year into account.

Table 1 summarises the methodological features, explana-

tions and limitations of the studies included.

3. Findings

Table 2 provides an overview of all findings according to coun-

try of study, population of interest and cancer site, expressed

in tendencies.

The all-cancer comparison of most studies showed in par-

ticular on average a lower cancer risk for first-generation mi-

grants from non-western countries in terms of incidence and

mortality, although there were some studies that did not re-

veal significant differences, sometimes obviously due to small

study cohorts. However, male subjects originating from West

Africa exhibited significantly elevated cancer mortality in two

studies from the United Kingdom.22,23 Ambiguous results

were attained for migrants from Eastern Europe: Many stud-

ies revealed advantageous risks, although in several cases

they were not significant.

Since all-cancer morbidity reflects a summary of site-spe-

cific results, the aim is to point out cancers with significantly

elevated or lowered risks among migrants and to investigate

these results according to migrant origin.

3.1. Migrants from Southern Europe

In 35% of all studies (13/37) included from five different coun-

tries, migrants from Southern Europe, mostly Turkey, were

investigated. According to these studies, all malignant neo-

plasms together tended to occur significantly less often in

this group compared with the general population of the host

country.

Significantly elevated risks for this migrant group could be

observed for cancers of the stomach, liver, lung among males

and thyroid gland. In addition, increased risks were reported

for Hodgkin’s disease and lymphomas. In contrast, signifi-

cantly lower risks were found for cancers of the oesophagus,

colorectum, lung among females, skin, breast, prostate and

testis and bladder.

3.2. Migrants from Eastern Europe

In 32% of studies from five countries (12/37) migrants came

from the Eastern part of Europe, mostly parts of the former

Soviet Union. Lower all-cancer morbidity and mortality were

confirmed by the majority of these studies.

The site-specific results were ambiguous, but strongly con-

curred on the elevated risks for stomach and lung cancer in

males, whereas consistently decreased risks could be ob-

served for breast cancer in females and melanoma.

3.3. Migrants from Africa

Migrants originating from the African continent had to be cat-

egorised into ‘Africa’ (if no subgroups were available), ‘North

Africa’, ‘West Africa’ and ‘East Africa’.

In 16% of studies from four countries (6/37) migrants from

Africa without further regional classifications were investi-

gated. However, only three studies covered all-cancer morbid-

ity which resulted in advantageous risks for migrants. The

most striking similarities in the study results could be ob-

served for liver cancer due to strongly elevated risks and colo-

rectal cancer as well as cancer of male and female genital

organs due to decreased risks.

North African migrants were studied in 12 studies (32%)

from five countries (Denmark, France, Sweden, Netherlands

and the UK). All-cancer morbidity was lower or not significant

in all studies. Elevated risks were observed for cancers of the

nasopharynx, liver, gallbladder and cervix uteri. Significantly

decreased risks were found for almost all other cancer sites,

especially for colorectal, lung and breast cancer as well as

melanoma.

Migrants from the western part of Africa represent an

exceptional group among migrants from non-industrialised

countries. Only 4 out of 37 studies (11%) from France and the

UK looked at this group but all of them presented quite detailed

results that allowed us to look at many possible parallels. All-

cancer mortality was significantly elevated among males

residing in the United Kingdom, but the opposite was the case

for males living in France. The studies coincide in increased

risks for cancers of the liver, pancreas and prostate as well as

Table 2 – Site-specific cancer occurrence in male and female migrants from different regions residing in selected Europeancountries.*,#

E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9 2655

2656 E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

lymphomas. Other cancer sites showed ambiguous results, for

example, significantly elevated mortality due to breast cancer

in the studies from the UK as opposed to study results from

France, which showed a significantly decreased risk among

West African women. This implies important regional risk

diversity in similar migrant groups across European countries

and is certainly an interesting subject for further research.

Another four studies from Sweden and the UK focussed on

East African migrants. The three British studies agreed on

lower all-cancer mortality in this group and revealed elevated

risks for cancer of the oral cavity and leukaemia. All other

cancer sites showed continuously decreased risks, most

remarkably for cancers of the colon and rectum, lung and

genital organs. The Swedish study yielded a significantly de-

creased risk of cancer of the cervix uteri in this migrant group.

3.4. Migrants from the Middle East

In 24% of the studies (9/37) migrants originating from the Mid-

dle East were investigated, investigating only few cancer sites.

All-cancer occurrence appeared to be significantly less fre-

quent in three studies. Moreover, decreased risks could be ob-

served for colorectal, lung, prostate, testis and breast cancer

in studies carried out in Denmark, the United Kingdom and

Sweden, where an increased risk of cancer of the thyroid

gland was also revealed.

3.5. Migrants from Asia

Many studies took migrants from Asia into account. With re-

gard to the vastness of this continent, it made sense to distin-

guish between Asia in general, mostly referring to China and

Vietnam, and South East Asia, which included India, Ceylon,

Bangladesh, Indonesia and sometimes Pakistan (depending

on the definition).

Cancer risks among migrants from Asia in general were

examined in 30% of the studies from six different European

countries (11/37), all of them exhibiting lower all-cancer mor-

tality and morbidity rates. Consistent findings of elevated

Fig. 1 – Cancer incidence in less and more developed regio

risks were found for cancers of the nasopharynx, stomach, li-

ver and endocrine glands as well as lymphomas. Parallel, de-

creased risks could in particular be observed for colorectal,

lung, breast and bladder cancer as well as for melanoma

and cancers of the cervix, ovary, prostate and testis.

Migrants from South East Asia showed surprisingly similar

results between the studies for many cancers. In total, 41% of

all studies included (15/37), performed in France, Sweden,

The Netherlands and the UK focused on this migrant group,

varying little in the definition of South East Asian countries.

All-cancer mortality and morbidity risks appeared to be con-

sistently lower in all studies that covered this general compar-

ison. Uniformly elevated risks were revealed for migrants with

cancers of the oral cavity, nasopharynx, liver, gallbladder and

thyroid gland. Moreover, a higher risk of lymphomas and leu-

kaemia was observed in several studies, whereas lowered risks

were found for stomach, colorectal, lung, breast, ovary, pros-

tate, testis, kidney and bladder cancer as well as melanoma.

3.6. Migrants from South and Central America

In 41% of all studies included in this overview (15/37) cancer

risks were determined for migrants coming from South and

Central American countries, most frequently Caribbean coun-

tries that used to be European colonies. All-cancer mortality

and morbidity risks were consistently lower for migrants

from this part of the world. Particularly elevated risks could

be observed for cancers of the nasopharynx, liver, cervix uteri,

prostate and lymphomas. In contrast, notably lowered risks

were revealed for cancers of the oesophagus, colon and rec-

tum, lung, breast, skin, ovary and bladder.

3.7. Second-generation migrants

Studies on cancer risk in second-generation migrants are still

scarce and were included in this overview for the sake of

completeness only. However, a convergence of risks towards

the rates of the host population as well as less extreme risks

was revealed by Hemminki and Li.12 In addition, studies

ns for males (a) and females (b) according to IARC 2002.

E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9 2657

assessing the effects of duration of residence or age at migra-

tion indicate an adaptation of rates, which also indicates a

change of risk over time, i.e. after migration. Investigating

cancer occurrence in second-generation migrants will be-

come more relevant in future, due to the increasing age and

size of this population group.

4. Discussion

Our findings suggest that migrants from non-western coun-

tries were more likely to develop cancers that are related to

infectious diseases, compared with the general population

of their industrialised host country. This is especially true

for cancers of the oral cavity, nasopharynx, stomach, liver,

gallbladder, cervix uteri, prostate and lymphomas. In con-

trast, almost all studies found lower risks for cancers that

are strongly related to a ‘western’ lifestyle (poor diet, physical

inactivity, reproductive factors, etc.), irrespective of the mi-

grant origin. This is in particular the case for colorectal cancer

and cancers of the pancreas, lung, breast, ovary, kidney and

bladder. Some elevated risks could also be explained partly

by important covariables such as socioeconomic status, espe-

cially for migrants originating from West Africa.

We also found that in most studies, migrants show cancer

risks that are in between the corresponding risk of the native

populations in their home and their host country. The major-

ity of the findings tend to be in accordance with the rates, vis-

ualised in Fig. 1.

Whereas all-cancer incidence in the more developed coun-

tries amounts to 314 [age-standardised rate (ASR(W)) per

100,000] among males and 228 among females, less well-

developed countries show an average of 159 for males and

129 for females.24

It can be observed that cancer sites with a comparatively

high incidence in less well-developed regions also exhibit a

high incidence for migrant populations from non-western

countries residing in industrialised countries. This applies

particularly for cancers of the liver, oesophagus, stomach

and nasopharynx among males and cervix, stomach, liver,

oesophagus and nasopharynx among females. In the same

manner, low incidences in less well-developed regions are re-

flected by low incidences among migrant groups originating

from these countries. This pattern could be confirmed in a re-

cent study by Zanetti and colleagues,25 who analysed cancer

incidence in North Africa.

Mortality data show a similar picture, although the differ-

ences are less clear, which is mainly attributable to disparities

in access to care and suboptimal communication on the

dilemmas of treatment.

Our findings also concur with those of others from non-

European countries and continents that host non-western

migrants. McCredie and colleagues26 for instance observed

lower cancer incidences for migrants from various non-wes-

tern origins in Australia and McDonald and Neily27 could con-

firm similar results for migrant women in the United States.

A close relationship with individual exposure experienced

during a life span could be confirmed for migrants of various

origins. In addition to individual factors and health behav-

iour, the causal roles of exposure in the home country, i.e. be-

fore migration, during migration itself and in the host

country, as well as the influence of social factors, certainly

represent key factors in carcinogenesis.

Exposure to risk factors and adaptation to changing envi-

ronments evolve over time and therefore cancer risk diversi-

fies with the duration of residence, new exposures and new

generations. Prospectively, a convergence of cancer risk (a

simultaneous decrease in cancers with high incidence in mi-

grants and an increase in those with a currently low inci-

dence) towards the level of the rates in the native

population of the host country can be expected over time

and across migrant generations.6,14,16,28

Of course there are limitations to the comparisons con-

ducted in this overview. Firstly, the definitions of the migrant

groups and the study populations varied among studies and

countries. Ethnicity proxies, such as ‘self-assigned ethnicity’

and name-based approaches, are in particular prone to mis-

classification bias, since a distinction between generations

or for example intercultural marriages is not possible. Sec-

ond, the comparability of studies is also limited with regard

to the size, composition and time window of the study popu-

lations. It is also important to note that in some studies pop-

ulation data from censuses or surveys were used (instead of

population-based registers), which is always a biased under-

estimate of the population at risk because as a rule only the

head of the household is considered.

Third, migrant origins may sometimes have been collected

in an inconsistent way, which was unavoidable in some cases

(e.g. the allocation of Pakistan or Turkey).

Fourth, studies investigating both mortality and morbidity

have been included, given the assumption of parallel effects,

although mortality is mainly driven by (access to) treatment

and the varying fatality rates of different cancers. Consequently,

different measures of association have been pooled and

compared on the basis of tendencies. The comparisons

therefore lack a magnitude and only provide a rough estimation

of risk disparities. Meta-analysis was not the aim of this

overview.

The healthy migrant effect could partly explain the advan-

tageous risks of migrants, but since effects seem to persist, its

influence is probably marginal. Several studies also discussed

the possible effects of the so-called salmon-bias, which

assumes that migrants tend to return to their roots when they

become ill. This is in most instances unlikely due to the fact

that health services and treatment are often better in the host

country and many migrants have already been joined by and

settled with their families.

This is to our knowledge the first direct comparison of

studies on cancer occurrence in migrant populations in

Europe. Despite the limitations mentioned above, broad

comparisons are feasible and will gain importance in the

future. Prospectively, a transnational study of cancer occur-

rence in migrant populations could surmount many of these

difficulties. This primarily concerns the definition of migrant

groups requiring close networking between countries. In

doing so, the results would be more reliable and the magni-

tude of the risk diversity could be studied in more detail. In

order to appreciate the change in risk after migration, a

comparison with data from the country of birth would be

ideal.

2658 E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9

Conflict of interest statement

None declared.

Acknowledgement

Melina Arnold’s work has partly been funded by EU SANCO

(MEHO project: Migrant and Ethnic Minorities Health Obser-

vatory; Contract number: 2005122).

R E F E R E N C E S

1. World Migration 2008: managing labour mobility in theevolving global economy; 2009.

2. Bhopal RS. Ethnicity, race, and health in multicultural societies:foundations for better epidemiology, public health, and healthcare. Oxford University Press; 2007.

3. Spallek J, Razum O. Health of migrants: deficiencies in thefield of prevention. Med Klin (Munich) 2007;102(6):451–6.

4. Omran AR. The epidemiologic transition. A theory of theepidemiology of population change. Milbank Mem Fund Q1971;49(4):509–38.

5. Marmot M. Changing places changing risks: the study ofmigrants. Public Health Rev 1993;21(3–4):185–95.

6. Razum O, Twardella D. Time travel with Oliver Twist – towardsan explanation for a paradoxically low mortality amongrecent immigrants. Trop Med Int Health 2002;7(1):4–10.

7. Razum O, Rohrmann S. The healthy migrant effect: role ofselection and late entry bias. Gesundheitswesen2002;64(2):82–8.

8. Parkin DM, Khlat M. Studies of cancer in migrants: rationaleand methodology. Eur J Cancer 1996;32A(5):761–71.

9. Zeeb H, Spallek J, Razum O. Epidemiological perspectives ofmigration research: the example of cancer. PsychotherPsychosom Med Psychol 2008;58(3–4):130–5.

10. Spallek J, Razum O. Erklarungsmodelle fur diegesundheitliche Situation von Migrantinnen und Migranten.In: Bauer U, Bittlingmayer UH, Richter M, editors. Healthinequalities: Determinanten und Mechanismen gesundheitlicherUngleichheit. Wiesbaden: Vs Verlag; 2008. p. 271–88.

11. Stronks K, Kulu-Glasgow I, Agyemang C. The utility of‘country of birth’ for the classification of ethnic groups inhealth research: the Dutch experience. Ethn Health2009;14(3):1–14.

12. Hemminki K, Li X. Cancer risks in second-generationimmigrants to Sweden. Int J Cancer 2002;99(2):229–37.

13. Myrup C, Westergaard T, Schnack T, et al. Testicular cancerrisk in first- and second-generation immigrants to Denmark. JNatl Cancer Inst 2008;100(1):41–7.

14. Stirbu I, Kunst AE, Vlems FA, et al. Cancer mortality ratesamong first and second generation migrants in theNetherlands: convergence toward the rates of the nativeDutch population. Int J Cancer 2006;119(11):2665–72.

15. Barker RM, Baker MR. Incidence of cancer in Bradford Asians.J Epidemiol Community Health 1990;44(2):125–9.

16. Spallek J, Arnold M, Hentschel S, Razum O. Cancerincidence rate ratios of Turkish immigrants in Hamburg,Germany: a registry based study. Cancer Epidemiol2009;33(6):413–8.

17. Jack RH, Davies EA, Moller H. Breast cancer incidence, stage,treatment and survival in ethnic groups in South EastEngland. Br J Cancer 2009;100(3):545–50.

18. Smith LK, Botha JL, Benghiat A, Steward WP. Latest trends incancer incidence among UK South Asians in Leicester. Br JCancer 2003;89(1):70–3.

19. Swerdlow AJ, Marmot MG, Grulich AE, Head J. Cancermortality in Indian and British ethnic immigrants from theIndian subcontinent to England and Wales. Br J Cancer1995;72(5):1312–9.

20. Winter H, Cheng KK, Cummins C, et al. Cancer incidence inthe south Asian population of England (1990–92). Br J Cancer1999;79(3–4):645–54.

21. Zeeb H, Razum O, Blettner M, Stegmaier C. Transition incancer patterns among Turks residing in Germany. Eur JCancer 2002;38(5):705–11.

22. Wild SH, Fischbacher CM, Brock A, Griffiths C, Bhopal R.Mortality from all cancers and lung, colorectal, breast andprostate cancer by country of birth in England and Wales,2001–2003. Br J Cancer 2006;94(7):1079–85.

23. Grulich AE, Swerdlow AJ, Head J, Marmot MG. Cancermortality in African and Caribbean migrants to England andWales. Br J Cancer 1992;66(5):905–11.

24. GLOBOCAN database database on the Internet. IARC; 2002.25. Zanetti R, Tazi MA, Rosso S. New data tells us more about

cancer incidence in North Africa. Eur J Cancer 2010;46(3):462–6.26. McCredie M, Coates MS, Ford JM. Cancer incidence in

migrants to New South Wales. Int J Cancer 1990;46(2):228–32.27. McDonald JT, Neily J. Race, immigrant status, and cancer

among women in the United States. J Immigr Minor Health2009. doi:10.1007/s10903-009-9268-1.

28. Harding S, Rosato M, Teyhan A. Trends in cancer mortalityamong migrants in England and Wales, 1979–2003. Eur JCancer 2009;45(12):2168–79.

29. Norredam M, Krasnik A, Pipper C, Keiding N. Differences instage of disease between migrant women and native Danishwomen diagnosed with cancer: results from a population-based cohort study. Eur J Cancer Prev 2008;17(3):185–90.

30. Norredam M, Krasnik A, Pipper C, Keiding N. Cancerincidence among 1st generation migrants compared to nativeDanes – a retrospective cohort study. Eur J Cancer2007;43(18):2717–21.

31. Bouchardy C, Parkin DM, Wanner P, Khlat M. Cancer mortalityamong north African migrants in France. Int J Epidemiol1996;25(1):5–13.

32. Bouchardy C, Wanner P, Parkin DM. Cancer mortality amongsub-Saharan African migrants in France. Cancer Causes Control1995;6(6):539–44.

33. Khlat M. Cancer in Mediterranean migrants-based on studiesin France and Australia. Cancer Causes Control1995;6(6):525–31.

34. Bouchardy C, Parkin DM, Khlat M. Cancer mortality amongChinese and South-East Asian migrants in France. Int J Cancer1994;58(5):638–43.

35. Winkler V, Ott JJ, Holleczek B, Stegmaier C, Becher H. Cancerprofile of migrants from the Former Soviet Union in Germany:incidence and mortality. Cancer Causes Control 2009.doi:10.1007/s10552-009-9381-4.

36. Ronellenfitsch U, Kyobutungi C, Ott JJ, et al. Stomachcancer mortality in two large cohorts of migrants from theFormer Soviet Union to Israel and Germany: are thereimplications for prevention? Eur J Gastroenterol Hepatol2009;21(4):409–16.

37. Ott JJ, Paltiel AM, Winkler V, Becher H. Chronic diseasemortality associated with infectious agents: a comparativecohort study of migrants from the Former Soviet Union inIsrael and Germany. BMC Public Health 2008;8:110.

38. Kyobutungi C, Ronellenfitsch U, Razum O, Becher H. Mortalityfrom cancer among ethnic German immigrants from theFormer Soviet Union, in Germany. Eur J Cancer2006;42(15):2577–84.

E U R O P E A N J O U R N A L O F C A N C E R 4 6 ( 2 0 1 0 ) 2 6 4 7 – 2 6 5 9 2659

39. Regidor E, de La Fuente L, Martinez D, Calle ME, Dominguez V.Heterogeneity in cause-specific mortality according tobirthplace in immigrant men residing in Madrid, Spain. AnnEpidemiol 2008;18(8):605–13.

40. Hemminki K, Mousavi SM, Brandt A, Ji J, Sundquist J. Liver andgallbladder cancer in immigrants to Sweden. Eur J Cancer2010;46(5):926–31.

41. Mousavi SM, Brandt A, Weires M, et al. Cancer incidenceamong Iranian immigrants in Sweden and Iranian residentscompared to the native Swedish population. Eur J Cancer2010;46(3):599–605.

42. Mousavi SM, Sundquist J, Hemminki K. Nasopharyngeal andhypopharyngeal carcinoma risk among immigrants inSweden. Int J Cancer 2010. doi:10.1002/ijc.25287.

43. Azerkan F, Zendehdel K, Tillgren P, Faxelid E, Sparen P. Risk ofcervical cancer among immigrants by age at immigration andfollow-up time in Sweden, from 1968 to 2004. Int J Cancer2008;123(11):2664–70.

44. Moradi T, Nordqvist T, Allebeck P, Galanti MR. Risk of thyroidcancer among Iranian immigrants in Sweden. Cancer CausesControl 2008;19(3):221–6.

45. Hemminki K, Li X, Czene K. Cancer risks in first-generationimmigrants to Sweden. Int J Cancer 2002;99(2):218–28.

46. Visser O, van Leeuwen FE. Cancer risk in first generationmigrants in North-Holland/Flevoland, The Netherlands, 1995–2004. Eur J Cancer 2007;43(5):901–8.

47. Bos V, Kunst AE, Keij-Deerenberg IM, Garssen J, MackenbachJP. Ethnic inequalities in age- and cause-specific mortality inThe Netherlands. Int J Epidemiol 2004;33(5):1112–9.

48. Visser O, van der Kooy K, van Peppen AM, Ory FG, vanLeeuwen FE. Breast cancer risk among first-generationmigrants in the Netherlands. Br J Cancer 2004;90(11):2135–7.

49. Visser O, Busquet EH, van Leeuwen FE, Aaronson NK, Ory FG.Incidence of cervical cancer in women in North-Holland bycountry of birth from 1988–1998. Ned Tijdschr Geneeskd2003;147(2):70–4.

50. Harding S, Rosato M. Cancer incidence among first generationScottish, Irish, West Indian and South Asian migrants livingin England and Wales. Ethn Health 1999;4(1–2):83–92.

51. Haworth EA, Soni Raleigh V, Balarajan R. Cirrhosis andprimary liver cancer amongst first generation migrants inEngland and Wales. Ethn Health 1999;4(1–2):93–9.

52. Swerdlow A. Mortality and cancer incidence in Vietnameserefugees in England and Wales: a follow-up study. Int JEpidemiol 1991;20(1):13–9.