WK 3 DIS EPID

Cancer Epidemiology 69 (2020) 101827

Available online 7 October 2020 1877-7821/© 2020 Elsevier Ltd. All rights reserved.

Socio-economic inequalities on cancer mortality in nine European areas: The effect of the last economic recession

Laia Palència a,b,c,*, Josep Ferrando a, Marc Marí-Dell’Olmo a,b,c, Mercè Gotsens a,c, Joana Morrison d, Dagmar Dzurova e, Michala Lustigova e, Claudia Costa f, Maica Rodríguez-Sanz a,b,c,g, Lucia Bosakova h,i, Paula Santana f, Carme Borrell a,b,c,g

a Agència de Salut Pública de Barcelona, Barcelona, Spain b CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain c Institut d’Investigació Biomèdica (IIB Sant Pau), Barcelona, Spain d Institute of Health Equity at the Research Department of Epidemiology and Public Health, University College London, London, United Kingdom e Department of Social Geography and Regional Development, Faculty of Science, Charles University, Prague, Czechia f Centre of Studies in Geography and Spatial Planning, University of Coimbra, Coimbra, Portugal g Universitat Pompeu Fabra, Barcelona, Spain h Department of Health Psychology and Research Methodology, Medical Faculty, P. J. Safarik University in Kosice, Kosice, Slovak Republic i Olomouc University Social Health Institute (OUSHI), Palacky University in Olomouc, Olomouc, Czech Republic

A R T I C L E I N F O

Keywords: Cancer Europe Socio-Economic inequalities Economic recession

A B S T R A C T

Background: The effect of inequalities aggravated by economic recessions in the mortality rates of certain diseases has been previously described. In this study, we analyzed the relationship between socio-economic deprivation and cancer mortality. We focused on lung, colon, prostate, and breast cancers in nine European urban areas over three periods: two before (2000–2003 and 2004–2008) and one after (2009–2014) the onset of the 2008 financial crisis. Methods: This is an ecological study of trends. The units of analysis were small areas within nine European urban areas. We used a composite deprivation index as a socio-economic indicator. As a mortality indicator, we used the smoothed standardized mortality ratio, calculated using the hierarchical Bayesian model proposed by Besag, York and Mollié. To analyze the evolution of socio-economic inequalities, we fitted an ecological regression model that included the socio-economic indicator, the period of time, and the interaction between these terms. Results: In men, socio-economic inequalities in all-cancer and lung cancer mortality were observed in most of the cities studied, but did not increase after the onset of the economic crisis. In women, only two cities (Stockholm and London) showed socio-economic inequalities in all-cancer and lung cancer mortality; there was also no increase in inequalities. Conclusions: Our results did not validate our hypothesis that inequalities increase in times of crisis. However, they emphasize the importance of socio-economic measurements for understanding mortality inequalities, and can be used to inform prevention strategies and help plan local health programs aimed at reducing health inequalities.

1. Introduction

Several studies have evaluated the relationship between economic crises and health, with some showing a direct relationship between crises and a general decline in population’s health, while others have claimed the opposite [1–4]. In the European Union, the overall reduc- tion in public investment, including health care, in times of crisis has been associated with an overall increase in mortality [5]. However, not

all studies support this assertion, with only some causes of mortality, such as suicide, increasing during these periods [6,7]. There is even evidence of a reduction of other types of mortality, for instance those caused by traffic injuries, in this case due to a decrease in the number of circulating vehicles [8,9].

Cancer is one of the main causes of death in the European population, specifically lung and colon cancer in both sexes, and prostate and breast cancer in men and women, respectively [10,11]. While there was a

* Corresponding author at: Agència de Salut Pública de Barcelona, Plaça Lesseps 1, 08023, Barcelona, Spain. E-mail addresses: [email protected] (L. Palència), [email protected] (M. Marí-Dell’Olmo).

Contents lists available at ScienceDirect

Cancer Epidemiology

journal homepage: www.elsevier.com/locate/canep

https://doi.org/10.1016/j.canep.2020.101827 Received 25 February 2020; Received in revised form 10 September 2020; Accepted 13 September 2020

Cancer Epidemiology 69 (2020) 101827

2

general decline in cancer mortality in the years preceding the crisis, the rate of this decline began to slow after the onset of the 2008 crisis [12, 13]. During economic crises, increased unemployment and decreased public investment has been linked to a rise in total cancer mortality, most noticeably in lung, colon, prostate, and breast cancers [14–18]. This rise in mortality could be counteracted by an increase in health services investment [15,19].

Economic crises, and their impact on health care, most negatively affect individuals from the most socio-economically disadvantaged areas and social classes, ethnic minorities, the unemployed, and those with lower levels of education [20–22]. However, few studies have analyzed the impact of economic crises on health inequalities, especially in urban areas, and their results have not always been consistent [20, 23–28].

The aim of this study was to analyze the effect of socio-economic inequalities on cancer mortality, specifically lung, colon, prostate, and breast cancers in nine European urban areas over three periods: before (two intervals: 2000− 2003 and 2004− 2008) and after (one interval: 2009− 2014) the onset of the 2008 financial crisis. Our hypotheses were that cancer mortality would be higher in the most disadvantaged socio- economic areas, and that these inequalities have increased after the onset of the economic crisis as the population living in these areas tends to suffer more from the negative effects of the crises.

2. Methods

2.1. Design, units of analyses, and study population

This study was part of a wider European project called Euro-healthy (http://www.euro-healthy.eu/), which aims to increase knowledge and resources related to policies promoting health and health equity across Europe, with a focus on metropolitan areas. We performed an ecological study of trends based on three study periods, two before the economic crisis (2000− 2003 and 2004− 2008), and one after the crisis (2009− 2014). The units of analysis were small areas in nine European urban areas: Athens metropolitan area, Barcelona, Berlin-Brandenburg Metropolitan Region, Brussels-Capital Region, Lisbon Metropolitan Area, Greater London, Prague, Turin, and Stockholm Metropolitan Area (see descriptions of the small areas in Table 1). We selected these areas because of the availability of data within the Euro-healthy project, which covers a broad range of geographical areas and socio-economic conditions. The study population consisted of individuals living in these areas during the three time periods.

2.2. Information sources

For most of the urban areas, mortality and population data were available for the three time periods (Table 1), and socio-economic in- dicators for 2001. We mainly obtained mortality data from mortality registers. We obtained population data stratified by age (five-year groups) and by sex from censuses or population registers. We obtained socio-economic data mainly from census records.

2.3. Variables and indicators

In this study we analyzed deaths due to (International Classification of Diseases, 10th revision, ICD10): all cancers (all neoplasms, C00 to D48); Lung Cancer (malignant neoplasms in the larynx, trachea, bron- chus or lung, C32 to C34); Colon Cancer (malignant neoplasms in the colon, rectum, anus, or anal canal, C18 to C21); Prostate Cancer (ma- lignant neoplasms in the prostate, C61) and Breast Cancer (malignant neoplasms in the female breast, C50).

As a mortality indicator, we computed the Standardized Mortality Ratio (SMR) for each small area of residence, and during each of the three time periods. To describe mortality for the whole city/metropol- itan area in each period, we calculated the mortality rate (MR) per Ta

bl e

1 D

es cr

ip tio

n of

th e

ni ne

E ur

op ea

n ur

ba n

ar ea

s: n

um be

r an

d ty

pe o

f s m

al l a

re as

, s tu

dy p

er io

d, to

ta l u

rb an

a re

a po

pu la

tio n,

a nd

m ed

ia n

si ze

o f t

he s

m al

l a re

as , f

or m

en a

nd w

om en

.

U rb

an a

re a

N ic

kn am

e N

um be

r of

s m

al l a

re as

Ty

pe o

f s m

al l a

re as

Ye ar

in te

rv al

s Po

pu la

tio n

M EN

W

O M

EN

1s t p

er io

d 2n

d pe

ri od

3r

d pe

ri od

To

ta l

M ed

ia n

To ta

l M

ed ia

n

A th

en s

m et

ro po

lit an

a re

a A

th en

s 40

m

un ic

ip al

iti es

20

00 −

20 03

20

04 −

20 08

20

09 −

20 13

1

57 7

17 2

29 7

45

1 71

0 44

6 32

1 63

Ba

rc el

on a

ci ty

Ba

rc el

on a

14 91

ce

ns us

tr ac

ts

20 00

− 20

03

20 04

− 20

08

20 09

− 20

13

69 7

56 3

45 7

79 6

49 7

51 7

Be rl

in -B

ra nd

en bu

rg M

et ro

po lit

an r

eg io

n Be

rl in

30

di

st ri

ct s

20 02

20

06

20 11

2

92 7

61 6

96 1

76

3 04

7 18

8 97

4 54

Br

us se

ls -c

ap ita

l r eg

io n

Br us

se ls

14

5 ne

ig hb

or ho

od s

20 01

− 20

03

20 04

− 20

08

20 09

− 20

11

46 4

36 4

4 00

4 50

5 67

3 4

28 8

Li sb

on m

et ro

po lit

an a

re a

Li sb

on

18 8

pa ri

sh es

20

00 −

20 03

20

04 −

20 08

20

09 −

20 12

1

27 5

81 3

5 43

7 1

38 6

31 4

5 83

5 G

re at

er L

on do

n Lo

nd on

98

3 ce

ns us

tr ac

ts

20 00

− 20

03

20 04

− 20

08

20 09

− 20

14

3 59

7 12

0 3

81 0

3 72

5 28

3 3

96 0

Pr ag

ue c

ity

Pr ag

ue

57

di st

ri ct

s 20

01 −

20 03

20

04 −

20 08

20

09 −

20 14

54

9 65

2 2

20 6

61 0

46 6

2 10

0 St

oc kh

ol m

m et

ro po

lit an

a re

a St

oc kh

ol m

12

99

ce ns

us tr

ac ts

20

01 −

20 03

20

04 −

20 08

20

09 −

20 11

89

7 48

7 56

0 93

6 97

7 59

9 Tu

ri n

ci ty

Tu

ri n

26 78

ce

ns us

tr ac

ts

20 00

− 20

03

20 04

− 20

08

20 09

− 20

13

42 5

78 2

12 9

46 5

98 7

14 2

L. Palència et al.

Cancer Epidemiology 69 (2020) 101827

3

100,000 inhabitants. We also used the indirectly standardized rate (ISR), calculated by multiplying the SMR for the city/metropolitan area by the crude rate in the standard population, which was considered to be the population of the 28 countries of the European Union (EU-28) in the year 2007 (representing approximately the mid-point of the entire studied interval).

We used a composite deprivation index as a socio-economic indicator for each small area. This index was the first component of a principal components analysis performed separately within each urban area (see supplementary material). It included: the rate of unemployment in people aged ≥ 16 years; the percentage of manual workers in people aged ≥ 16 years; the percentage of people aged 25–64 years who had primary education as the highest of education completed (International Standard Classification of Education [ISCED] level 0 and 1, except for London where we included ISCED levels 0, 1, and 2); and the percentage of people aged 25–64 years who had a university degree (ISCED 5 and 6) [29]. There were no data for the percentage of manual workers in Stockholm; therefore, we constructed the index using the other three available indicators.

2.4. Data analysis

The SMR depends on population size because its variance is inversely proportional to the expected values. Therefore, areas with low pop- ulations tend to have more variable estimates. To smooth the SMR for

each small area, we used the hierarchical Bayesian model proposed by Besag et al. [30], which accounts for two types of random effects, spatial and heterogeneous. The spatial effect considers the spatial structure of the data, while the heterogeneous effect deals with non-structural (non-spatial) variability. We used the following model to estimate the smoothed SMR (sSMR) for each sex, cause of death, and period:

Oi ~ Poisson (Eiθi)

log(θi) = α + Si + Hi, (model 1)

where, for each area i, Oi is the number of observed cases, Ei the ex- pected number of cases, θi the sSMR with respect to the European population, Si the spatial effect, and Hi the heterogeneous effect. We calculated the expected number of cases by indirect standardization, taking as reference the mortality rates of the EU-28 in 2007. We also considered age (in 5 year groups) and sex.

We used septile maps to represent the geographical distribution of the sSMR, the socio-economic indicators, and the composite deprivation indicator.

To analyze trends in socio-economic inequalities, we fitted an ecological regression model that includes the composite deprivation index (D), the period (through two dummy variables P2 and P3), and a term for interaction between these variables:

Oit ~ Poisson (Eitθit)

Table 2 Number of deaths, mortality rate, and Indirectly Standardized Mortality Rate per 100 000 inhabitants in for each study period: MEN.

ALL-CANCER First period Second period Third period

Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR

Athens 16 217 258.71 337.49 20 959 276.34 334.4 22 912 310.04 345.41 Barcelona 9 962 345.85 363.79 12 827 337.84 358.66 12 460 323.74 331.47 Berlin 8 099 276.64 370.6 8 322 275.23 333.03 9 131 311.59 324.96 Brussels 2 830 200 275.57 6 058 240.33 343.4 3 630 223.15 329.52 Lisbon 14 423 282.45 368.51 18 958 293.55 356.33 16 558 312.08 351.86 London 30 632 211.71 342.46 36 245 193.54 315.88 43 748 178.95 289.84 Prague 5 487 331.14 407.16 9 143 317.32 381.65 10 115 277.74 325.82 Stockholm 4 841 178.55 240.36 8 233 174.82 230.54 4 899 168.88 221.07 Turin 6 364 372.54 374.95 8 203 383.94 361.92 7 983 369.88 326.46 LUNG CANCER First period Second period Third period Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 5 352 85.38 105.16 6 863 90.49 104.66 7 394 100.06 108.06 Barcelona 2 932 101.79 102.52 3 739 98.48 101.76 3 559 92.47 94.15 Berlin 2 287 78.12 94.12 2 286 75.6 83.4 2 510 85.65 83.34 Brussels 822 58.09 77.09 1 829 72.56 100.5 1 104 67.87 97.88 Lisbon 3 223 63.12 76.14 4 363 67.56 76.79 3 699 69.72 74.61 London 8 120 56.12 87.31 9 122 48.71 76.67 10 406 42.56 66.59 Prague 1 449 87.45 100.21 2 296 79.69 90.27 2 445 67.14 75.1 Stockholm 1 041 38.39 49.48 1 765 37.48 47.23 1 014 34.95 43.46 Turin 2 006 117.43 110.75 2 470 115.61 103.97 2 317 107.35 93.07 COLON CANCER First period Second period Third period Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 1 180 18.82 25.58 1 755 23.14 28.88 2 031 27.48 31.22 Barcelona 1 224 42.49 45.54 1 659 43.7 46.95 1 635 42.48 43.63 Berlin 934 31.9 46.04 912 30.16 38.65 895 30.54 32.95 Brussels 249 17.6 25.04 612 24.28 35.87 389 23.91 36.52 Lisbon 2 072 40.58 55.78 2 831 43.84 55.41 2 406 45.35 52.61 London 2 980 20.6 34.81 3 442 18.38 31.39 4 265 17.45 29.57 Prague 876 52.87 68.25 1 300 45.12 56.55 1 312 36.03 43.66 Stockholm 620 22.87 32.03 1 055 22.4 30.72 654 22.54 30.72 Turin 712 41.68 43.16 960 44.93 43.06 937 43.41 38.32 PROSTATE CANCER First period Second period Third period Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 1539 24.55 36.76 2027 26.73 35.8 2074 28.07 33.28 Barcelona 767 26.63 29.94 958 25.23 27.89 967 25.12 25.94 Berlin 804 27.46 48.22 764 25.27 37.56 889 30.34 35.7 Brussels 312 22.05 33.54 577 22.89 36.15 347 21.33 34.76 Lisbon 1716 33.6 52.44 1995 30.89 43.07 1794 33.81 42.06 London 3546 24.51 45.33 4461 23.82 44.56 5371 21.97 40.8 Prague 534 32.23 46.51 911 31.62 43.57 1084 29.77 39.09 Stockholm 1188 43.82 66.72 2034 43.19 64.59 1092 37.64 56.34 Turin 502 29.39 33 629 29.44 29.52 624 28.91 25.66

L. Palència et al.

Cancer Epidemiology 69 (2020) 101827

4

log(θit) = α + β1Di + β2P2t + β3P3t + β4P2tDi + β5P3tDi + Sit + Hit(model 2)

where, for each area i and period t (t = 1, 2 and 3 for the first, second and third periods, respectively), Oit is the observed number of cases, Eit the expected number of cases, θit the sSMR with respect to the European population, Sit the spatial effect, and Hit the heterogeneous effect. Finally, P2t and P3t took the following values: Pjt = 1 if j = t, and 0 otherwise, where j = 2 or 3. We calculated the expected number of cases as in the previous model. We computed interaction terms in model 2 to evaluate changes between periods in the relationship between the socio-economic deprivation index and mortality.

In the two models, the spatial effect was assigned an intrinsic con- ditional autoregressive prior distribution (ICAR), which assumes that the expected value of each area is equal to the mean of the spatial effect of the adjacent areas, and has a variance of σs[2]. The heterogeneous effect was represented using independent normal distributions with mean of 0 and variance of σh[2]. A uniform distribution U(0,∞) was assigned to the standard deviations σs and σh. A normal vague prior distribution was assigned to the parameters α, β1, β2, β3, β4, and β5.

Since the composite deprivation index is dimensionless and arbi- trarily fixed, we calculated the Relative risk (RR), which compares the suicide mortality of the 95th percentile of socio-economic deprivation (severe deprivation) to that of the 5th percentile (low deprivation). We obtained RR estimates based on the mean of their posterior distribution,

along with their corresponding 95% credible intervals. We performed all analyses using R software version 3.5.0 [31] and

the INLA package [32].

3. Results

The largest of the studied areas by population was Greater London (7 322 403 inhabitants) and the smallest was Turin (891 769 inhabitants) (see Table 1). Berlin-Brandenburg has the smallest number of areas (30 areas) and Turin has the highest (2678 census tracts). Most small areas have a median population of fewer than 12 000 people except for the metropolitan areas of Athens and Berlin (where median small area size is ~60 000 and ~200 000, respectively).

Tables 2 and 3 show the number of deaths and the crude and indirect standardized mortality rates for cancers for men and women, respec- tively, during the three study periods. In men, cancer mortality, both for all cancers and for each type (lung, colon, and prostate cancer), tended to decrease over the three time periods in all cities and metropolitan areas, except for Athens, which experienced a slight increase in mor- tality due to lung, colon, and all cancers. In women, we observed a decrease in mortality rates during the three periods for colon, breast, and all cancers. However, mortality due to lung cancer increases in most cities and metropolitan areas, except in London and Prague.

Fig. 1 shows the association between the socio-economic deprivation index and cancer mortality in men. There were socio-economic

Table 3 Number of deaths, mortality rate, and Indirectly Standardized Mortality Rate per 100 000 inhabitants in for each study period: WOMEN.

ALL-CANCER First period Second period Third period

Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 11 957 175.8 191.38 15 852 191.85 191.33 17 133 211.51 191.26 Barcelona 7 006 215.25 167.44 8 934 212.09 165.99 9 222 216.73 167.19 Berlin 7 637 250.62 238.29 7 292 234.06 213.07 7 440 243.62 203.39 Brussels 2 755 179.42 175.25 5 848 216.84 222.75 3 450 201.12 216.53 Lisbon 10 514 189.11 193.88 13 467 189.42 182.91 11 929 202.12 183.83 London 29 064 194.44 248.28 33 918 175.48 234.27 40 218 160.71 220.55 Prague 5 226 285.48 255.73 8 656 278.84 245.48 9 577 248.43 218.73 Stockholm 4 432 156.87 165.21 7 322 150.31 159.34 4 319 145.18 157.01 Turin 5 361 286.82 218.86 6 490 276.84 202.91 6 532 275.06 192.19 LUNG CANCER First period Second period Third period Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 1 408 20.7 21.57 1 988 24.06 23.45 2 352 29.04 26.27 Barcelona 522 16.04 12.66 779 18.49 14.91 946 22.23 18.01 Berlin 920 30.19 28.04 1 004 32.23 28.8 1 214 39.75 32.91 Brussels 343 22.34 22.48 867 32.15 34.2 529 30.84 34.58 Lisbon 704 12.66 12.55 981 13.8 13.05 1 012 17.15 15.48 London 5 031 33.66 43.48 6 387 33.04 44.43 7 882 31.5 43.25 Prague 756 41.3 36.73 1 342 43.23 38.07 1 515 39.3 34.87 Stockholm 887 31.39 33.64 1 573 32.29 34.69 971 32.64 35.32 Turin 592 31.67 23.99 774 33.02 24.29 854 35.96 25.97 COLON CANCER First period Second period Third period Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 1 167 17.16 19.09 1 544 18.69 18.8 1 676 20.69 18.6 Barcelona 1 061 32.6 24.79 1 353 32.12 24.45 1 392 32.71 24.39 Berlin 1 084 35.57 34.06 917 29.43 26.84 805 26.36 21.84 Brussels 356 23.18 22.19 695 25.77 25.98 407 23.73 25.1 Lisbon 1 628 29.28 30.46 2 122 29.85 28.98 1 799 30.48 27.63 London 2 772 18.55 23.76 3 366 17.41 23.47 3 775 15.09 21.04 Prague 741 40.48 36.21 1 075 34.63 30.32 1 049 27.21 23.73 Stockholm 612 21.66 22.67 1 051 21.57 22.79 589 19.8 21.47 Turin 683 36.54 27.55 842 35.92 25.81 747 31.46 21.28 BREAST CANCER First period Second period Third period Urban area Deaths MR ISMR Deaths MR ISMR Deaths MR ISMR Athens 2 297 33.77 36.97 3 032 36.7 36.93 3 252 40.15 36.81 Barcelona 1 159 35.61 28.81 1 387 32.93 26.68 1 367 32.13 25.36 Berlin 1 179 38.69 37.05 1 157 37.14 34.27 1 254 41.06 35.14 Brussels 532 34.65 34.64 1 065 39.49 41.07 678 39.52 42.63 Lisbon 1 999 35.95 37.16 2 400 33.76 32.98 2 160 36.6 33.8 London 5 269 35.25 44.82 5 786 29.94 39.43 6 527 26.08 34.98 Prague 782 42.72 38.71 1 343 43.26 38.44 1 375 35.67 31.56 Stockholm 814 28.81 30.31 1 335 27.4 28.8 776 26.09 27.95 Turin 943 50.45 39.76 1 130 48.2 36.85 1 084 45.65 33.06

L. Palència et al.

Cancer Epidemiology 69 (2020) 101827

5

inequalities in the mortality rates for all cancers in all the studied pe- riods, with higher mortality in the most disadvantaged areas in all cities, except in Prague, Lisbon, and Berlin (which almost reached statistical significance). We obtained similar results for lung cancer mortality but not for colon and prostate cancer, for which no significant inequalities in mortality were observed (except in Stockholm and London for colon cancer mortality). However, comparing cancer mortality before and after the onset of the economic crisis, we observed no increase in in- equalities in any of the studied cities or metropolitan areas.

Fig. 2 shows the association between the socio-economic deprivation index and cancer mortality in women. There were no socio-economic inequalities in mortality due to all cancers in the small areas of most cities and metropolitan areas, except for London and Stockholm during the second and third periods. For lung cancer, Stockholm, London, and Brussels showed socio-economic inequalities in mortality during the third period, whilst in Lisbon mortality was higher in the least advan- taged areas during all the periods. There were no socio-economic in- equalities in mortality rates for the remaining cancer types, with two exceptions: risk of colon cancer was positively associated with socio- economic deprivation during the second period in Athens and nega- tively associated during the same period in Lisbon. Risk of breast cancer was also negatively associated with socio-economic deprivation in Lis- bon during the third period and in Barcelona during the first and third periods. Regarding trends in socio-economic inequalities, only Stock- holm and Lisbon showed changing trends during the second period, the former showing a significant increase and the latter a significant decrease in inequalities in all-cancer mortality. However, no changing trends were found during the third period, which corresponds to the interval following the onset of the economic crisis.

4. Discussion

For most of the cities analyzed in our study and for both sexes, we observed a decrease over time in mortality due to all cancers and all types of cancer studied, except lung cancer in women, which increased. In men, socio-economic inequalities in all-cancer and lung cancer mor- tality were observed in most of the cities studied, but did not increase after the onset of the economic crisis. In women, only two cities (Stockholm and London) showed socio-economic inequalities in all- cancer and lung cancer mortalities. Also, there was no increase in in- equalities in cancer mortality rates after the onset of the economic crisis.

In the context of the European Union, in both sexes there has been a general downward trend in mortality due to all cancers and most types of cancer, with the exception of lung cancer in women [33–38]. These trends are consistent with those observed in our study.

All-cancer mortality continued to decline after the onset of the 2008 economic crisis, including in countries that were strongly affected by the crisis [18,26]. However, in some cases the rate of this decline slowed due to the economic recession [13]. One explanation for this continuous decrease in mortality is that, despite economic cuts implemented in response to the crisis, countries with public health systems provide ac- cess to it for the entire population, which would have a dampening effect on the negative consequences of the health cuts [15].

Several studies have shown socio-economic inequalities in lung cancer mortality: mortality rates are higher in men and women with lower socio-economic status who are living in the most deprived areas. These inequalities are more striking in men than in women, likely due to exposure to various risk factors such as tobacco consumption [39,40]. Stress and lack of resources due to unemployment, which lead to smoke as a coping strategy, might explain the higher levels of

Fig. 1. Association between the socio-economic deprivation index and mortality, relative risk (RR) and 95% credible intervals (CI) for MEN in nine urban areas (Note: * indicates that RR2 is statistically significantly different from RR1 and + means that RR3 is statistically significantly different from RR2).

L. Palència et al.

Cancer Epidemiology 69 (2020) 101827

6

respiratory-related deaths in these populations [41]. The differences in global cancer mortality between men and women could be due to traditional gender roles, according to which men present riskier be- haviors than women [42].

Patients with lower socio-economic levels have greater difficulty in accessing lung cancer treatments [43], and in countries worst affected by the economic crisis, the prognosis was worse for these patients [44]. These results are consistent with those obtained in our study, in that lung cancer mortality was higher among men living in the most disad- vantaged socio-economic areas in most cities studied, and among women in some cities. However, in our study, these inequalities in mortality did not increase after the onset of the economic crisis.

Past socio-economic differences in breast cancer mortality have been decreasing due to a general improvement in health-related behaviors and access to breast cancer screening programs and health services [45, 46]. These trends are also reflected in our study, in that women with breast cancer who are living in more disadvantaged areas did not have higher mortality than those living elsewhere. Some studies also sug- gested that, in times of economic crisis, more disadvantaged breast cancer patients show a decrease in survival due to complications in accessing health services [14,47,48]. According to our results, in- equalities in breast cancer mortality did not increase after the onset of the economic crisis.

Previous studies report socio-economic inequalities in colon and prostate cancer mortality: colon cancer patients with lower socio- economic status or who live in more deprived urban areas have higher mortality [48–51]. Similar results have been found for prostate cancer [18,52–54]. There is also evidence of an increase in inequalities in mortality due to both colon and prostate cancer after the economic crisis [16,17]. Our results showed the opposite: mortality for colon and

prostate cancers were not higher among patients living in poorer urban areas, and inequality did not increase after the onset of the crisis.

As mentioned above, inequalities in cancer mortality could be due to differences in population exposure to various risk factors (e.g. tobacco), or to varying degrees of impediment to accessing health system services, such as if more vulnerable people have greater difficulty in accessing diagnostic tests and certain treatments. Since the cities and metropolitan areas studied generally had health systems with a strong public component [55], this may have contributed to the fact that we did not detect an increase in inequalities in cancer mortality after the beginning of the crisis.

One of the limitations of our study is that we analyzed only a six-year period post-crisis, which may not have been long enough to detect changes in cancer mortality. Given the latency period of the cancers studied (the latency time period between cause and cancer is often more than 10 years), the effects of the crisis may appear after a longer period of time. Therefore, in the coming years it would be advisable to continue monitoring mortality trends as well as trends in socio-economic markers. In addition, we used cancer mortality data, and this kind of data aggregate incidence and survival rates without being able to distinguish them. This can represent a limitation in interpreting the trends since incidence and lethality have different temporal trends for most cancers. However, socioeconomic inequalities in terms of which geographic areas have higher rates are likely to be similar for both outcomes. Furthermore, we only considered economic data for one year, which therefore assumes that the economic conditions were constant for the entire study period. Although socio-economic indicators may have changed, the ranking of the areas in terms of deprivation did not change much and we were not using the value itself but only the distribution. The sizes of the small areas differed between the urban areas studied.

Fig. 2. Association between the socio-economic deprivation index and mortality, relative risk (RR) and 95% credible intervals (CI) for WOMEN in nine urban areas (Note: * indicates that RR2 is statistically significantly different than RR1 and + means that RR3 is statistically significantly different than RR2).

L. Palència et al.

Cancer Epidemiology 69 (2020) 101827

7

Smaller areas are more homogeneous and the possibility of observing higher effects is higher, which may partly explain the inconsistency of some of our results.

5. Conclusions

There were inequalities in mortality among men: those living in the most deprived areas had higher mortality for all of the cancers studied, and for lung cancer in particular. In women, inequalities in cancer mortality were less common. Finally, we found that inequalities in mortality did not increase after the onset of the crisis, so our results did not validate our hypothesis of increasing inequalities. This is very likely due to the short period of time studied after the crisis. However, our results emphasize the importance of socio-economic measurements for understanding mortality inequalities, and can be used to inform pre- vention strategies and help plan local health programs aimed at reducing health inequalities. Future studies should study the role of cancer screening programs and other health systems characteristics on cancer inequalities.

Authors’ contributions

LP, JF, MM, MG and CB designed the paper, analyzed and interpreted the data; JM, DD, ML, CC, MR, LB and PS provided and interpreted the data; LP and JF wrote the first version of the manuscript; All authors revised it critically; All authors read and approved the final manuscript.

Funding

This study is a part of the EURO-HEALTHY project (Shaping EU- ROpean policies to promote HEALTH equity) and has received funding from the European Union’s Horizon 2020 research and innovation programme (Grant Agreement No 643398).

CRediT authorship contribution statement

Laia Palència: Conceptualization, Formal analysis, Writing - orig- inal draft. Josep Ferrando: Conceptualization, Writing - original draft. Marc Marí-Dell’Olmo: Conceptualization, Writing - review & editing. Mercè Gotsens: Methodology, Writing - review & editing. Joana Morrison: Writing - review & editing. Dagmar Dzurova: Writing - re- view & editing. Michala Lustigova: Writing - review & editing. Claudia Costa: Writing - review & editing. Maica Rodríguez-Sanz: Writing - review & editing. Lucia Bosakova: Writing - review & editing. Paula Santana: Writing - review & editing. Carme Borrell: Conceptualization, Writing - review & editing.

Declaration of Competing Interest

None.

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.canep.2020.101827.

References

[1] M.E. Falagas, E.K. Vouloumanou, M.N. Mavros, D.E. Karageorgopoulos, Economic crises and mortality: a review of the literature, Int. J. Clin. Pract. 63 (8) (2009) 1128–1135.

[2] C.E. Margerison-zilko, R. Catalano, S. Goldman-mellor, K. Saxton, C.E. Margerison- zilko, M. Subbaraman, et al., The Health Effects of Economic Decline The Health Effects of Economic Decline., 2016. March 2010.

[3] G.L. Quaglio, T. Karapiperis, L. Van Woensel, E. Arnold, D. McDaid, Austerity and Health in Europe. Health Policy (New York), 2013, pp. 13–19, https://doi.org/ 10.1016/j.healthpol.2013.09.005, 113(1–2)Available from:.

[4] G. Quaglio, T. Karapiperis, L. Van Woensel, E. Arnold, D. McDaid, C. Sargent, et al., Austerity and its implications for immigrant health in France, Soc. Sci. Med. 187 (1–2) (2017 Nov) 259–267, https://doi.org/10.1016/j.socscimed.2017.05.007 [cited 2016 Oct 21]Available from:.

[5] S. Budhdeo, J. Watkins, R. Atun, C. Williams, T. Zeltner, M. Maruthappu, Changes in government spending on healthcare and population mortality in the European union, 1995–2010: a cross-sectional ecological study, J. R. Soc. Med. 108 (12) (2015) 490–498.

[6] C. Coope, D. Gunnell, W. Hollingworth, K. Hawton, N. Kapur, V. Fearn, et al., Suicide and the 2008 economic recession: who is most at risk? Trends in suicide rates in England and Wales 2001-2011, Soc. Sci. Med. (117) (2014) 76–85, https:// doi.org/10.1016/j.socscimed.2014.07.024. Available from:.

[7] C. Haw, K. Hawton, D. Gunnell, S. Platt, Economic recession and suicidal behaviour: possible mechanisms and ameliorating factors, Int. J. Soc. Psychiatry 61 (1) (2015) 73–81.

[8] A. Baumbach, G. Gulis, Impact of financial crisis on selected health outcomes in Europe, Eur. J. Public Health 24 (3) (2014) 399–403.

[9] M. Karanikolos, P. Heino, M. McKee, D. Stuckler, H. Legido-Quigley, Effects of the global financial crisis on health in high-income OECD countries: a narrative review, Int. J. Heal. Serv. 46 (2) (2016) 208–240.

[10] C.C. Huisman, L.G.A. Bonneux, Health Statistics - Atlas on Mortality in the European Union, edition [Internet]. Eurostat statistical books. 2009. 205 p. Available from:, 2009 http://www.narcis.nl/publication/RecordID/oai%3Apure. knaw.nl%3Apublications%2Fe69db96f-0068-406f-a268-d601d4ecc424.

[11] J. Ferlay, M. Colombet, I. Soerjomataram, T. Dyba, G. Randi, M. Bettio, et al., Cancer incidence and mortality patterns in Europe: estimates for 40 countries and 25 major cancers in 2018, Eur. J. Cancer (103) (2018) 356–387. Available from: www.sciencedirect.com.

[12] N. Vrachnis, N. Vlachadis, N. Salakos, M. Vlachadi, Z. Iliodromiti, Cancer mortality in Greece during the financial crisis, Acta Oncol. (Madr). 54 (2) (2015) 287–288.

[13] J. Ferrando, L. Palència, M. Gotsens, V. Puig-Barrachina, M. Marí-Dell’Olmo, M. Rodríguez-Sanz, et al., Trends in cancer mortality in Spain: the influence of the financial crisis, Gac. Sanit. (2019).

[14] M. Maruthappu, J.A. Watkins, M. Waqar, C. Williams, R. Ali, R. Atun, et al., Unemployment, public-sector health-care spending and breast cancer mortality in the European Union: 1990-2009, Eur. J. Public Health 25 (2) (2015) 330–335.

[15] M. Maruthappu, J. Watkins, A.M. Noor, C. Williams, R. Ali, R. Sullivan, et al., Economic downturns, universal health coverage, and cancer mortality in high- income and middle-income countries, 1990-2010: a longitudinal analysis, Lancet 6736 (16) (2016) 1–12.

[16] M. Maruthappu, J. Watkins, A. Taylor, C. Williams, R. Ali, T. Zeltner, et al., Unemployment and prostate cancer mortality in the OECD, 1990-2009, Ecancermedicalscience 9 (2015) 1–13.

[17] M. Maruthappu, R.A. Watson, J. Watkins, C. Williams, T. Zeltner, O. Faiz, et al., Unemployment, public-sector healthcare expenditure and colorectal cancer mortality in the European Union: 1990–2009, Int. J. Public Health 61 (1) (2016) 119–130.

[18] K. Shafique, D.S. Morrison, Socio-economic inequalities in survival of patients with prostate Cancer: role of age and Gleason grade at diagnosis, PLoS One 8 (2) (2013).

[19] F. Ades, C. Senterre, E. de Azambuja, R. Sullivan, R. Popescu, F. Parent, et al., Discrepancies in cancer incidence and mortality and its relationship to health expenditure in the 27 European Union member states, Ann. Oncol. 24 (11) (2013) 2897–2902.

[20] A. Bacigalupe, A. Escolar-Pujolar, The impact of economic crises on social inequalities in health: What do we know so far? Int. J. Equity Health 13 (1) (2014) 1–6.

[21] A. Suess, I. Perez, A.R. Azarola, The right of access to health care for undocumented migrants : a revision of comparative analysis in the European context, Eur. J. Public Health (2015) 712–720, table 1.

[22] M. Gotsens, D. Malmusi, N. Villarroel, C. Vives-Cases, I. Garcia-Subirats, C. Hernando, et al., Health inequality between immigrants and natives in Spain: the loss of the healthy immigrant effect in times of economic crisis, Eur. J. Public Health 25 (6) (2015) 923–929.

[23] L. Maynou, M. Saez, G. Lopez-Casasnovas, Has the economic crisis widened the intraurban socioeconomic inequalities in mortality? The case of Barcelona, Spain, J. Epidemiol. Commun. Health. 70 (2) (2014) 114–124.

[24] L. Maynou, M. Saez, Economic crisis and health inequalities: evidence from the European Union, Int. J. Equity Health 15 (1) (2016) 1–11, https://doi.org/ 10.1186/s12939-016-0425-6. Available from:.

[25] R. Hoffmann, G. Borsboom, M. Saez, M.M. Dell’olmo, B. Burström, D. Corman, et al., Social differences in avoidable mortality between small areas of 15 European cities: an ecological study [Internet]. Vol. 13, Int. J. Health Geogr. (2014). Available from: http://www.ij-healthgeographics.com/content/13/1/8.

[26] E. Regidor, F. Vallejo, J.A.T. Granados, F.J. Viciana-Fernández, L. de la Fuente, G. Barrio, Mortality decrease according to socioeconomic groups during the economic crisis in Spain: a cohort study of 36 million people, Lancet 388 (10060) (2016) 2642–2652, https://doi.org/10.1016/S0140-6736(16)30446-9. Available from:.

[27] H. Nogueira, What is happening to health in the economic downturn? A view of the Lisbon Metropolitan Area, Portugal. Ann Hum Biol. 43 (2) (2016) 164–168.

[28] C. Borrell, L. Palència, M. Marí Dell’Olmo, J. Morrisson, P. Deboosere, M. Gotsens, et al., Socioeconomic inequalities in suicide mortality in European urban areas before and during the economic recession, Eur. J. Public Health (2019). Aug 13 [cited 2019 Aug 22]; Available from: https://academic.oup.com/eurpub/advance- article/doi/10.1093/eurpub/ckz125/5549577.

L. Palència et al.

Cancer Epidemiology 69 (2020) 101827

8

[29] International Standard Classification of Education (ISCED) 1997. Vol. http://www, UNESCO.

[30] J. Besag, J. York, A. Mollié, Bayesian image restoration, with two applications in spatial statistics, AnnInstStatistMath. 43 (1991) 1–59.

[31] R core Team, R: a Language and Environment for Statistical Computing [Internet], Available from:, R Foundation for Statistical Computing, Vienna, Austria, 2018 http://www.r-project.org/.

[32] H. Rue, S. Martino, INLA: Functions Which Allow to Perform a Full Bayesian Analysis of Structured Additive Models Using Integrated Nested Laplace Approximaxion, 2009.

[33] H.E. Karim-Kos, E. de Vries, I. Soerjomataram, V. Lemmens, S. Siesling, J.W. W. Coebergh, Recent trends of cancer in Europe: a combined approach of incidence, survival and mortality for 17 cancer sites since the 1990s, Eur. J. Cancer 44 (10) (2008) 1345–1389.

[34] J. Lortet-Tieulent, E. Renteria, L. Sharp, E. Weiderpass, H. Comber, P. Baas, et al., Convergence of decreasing male and increasing female incidence rates in major tobacco-related cancers in Europe in 1988-2010, Eur. J. Cancer 51 (9) (2015) 1144–1163, https://doi.org/10.1016/j.ejca.2013.10.014. Available from:.

[35] D.A. Ouakrim, C. Pizot, M. Boniol, M. Malvezzi, M. Boniol, E. Negri, et al., Trends in colorectal cancer mortality in Europe: retrospective analysis of the WHO mortality database, BMJ. 351 (2015) 1–10.

[36] D. Hashim, P. Boffetta, C. La Vecchia, M. Rota, P. Bertuccio, M. Malvezzi, et al., The global decrease in cancer mortality: trends and disparities, Ann. Oncol. 27 (5) (2016) 926–933.

[37] M.C.S. Wong, X.Q. Lao, K.F. Ho, W.B. Goggins, S.L.A. Tse, Incidence and mortality of lung cancer: global trends and association with socioeconomic status, Sci. Rep. 7 (1) (2017) 1–9, https://doi.org/10.1038/s41598-017-14513-7. Available from:.

[38] M. Malvezzi, G. Carioli, P. Bertuccio, P. Boffetta, F. Levi, C. La Vecchia, et al., European cancer mortality predictions for the year 2019 with focus on breast cancer, Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 30 (5) (2019) 781–787.

[39] G. Alicandro, G. Sebastiani, P. Bertuccio, N. Zengarini, G. Costa, C. La Vecchia, et al., The main causes of death contributing to absolute and relative socio-economic inequality in Italy, Public Health (2018).

[40] J.H.A. Van der Heyden, M.M. Schaap, A.E. Kunst, S. Esnaola, C. Borrell, B. Cox, et al., Socioeconomic inequalities in lung cancer mortality in 16 European populations, Lung Cancer. 63 (3) (2009) 322–330.

[41] L. Bosakova, K. Rosicova, D. Filakovska Bobakova, M. Rosic, D. Dzurova, H. Pikhart, et al., Mortality in the Visegrad countries from the perspective of socioeconomic inequalities, Int. J. Public Health 64 (3) (2019) 365–376. April 1 [cited 2020 Feb 11]Available from: http://www.ncbi.nlm.nih.gov/pubmed /30535783.

[42] C. Borrell, M. Rodríguez Sanz, M.I. Pasarín, M. Marí Dell’olmo, L. Palència, M. Gotsens, et al., Socioeconomic inequalities in mortality in 16 European cities, Scand. J. Public Health 42 (3) (2014) 245–254.

[43] L.F. Forrest, J. Adams, H. Wareham, G. Rubin, M. White, Socioeconomic inequalities in lung Cancer treatment: systematic review and meta-analysis, PLoS Med. 10 (2) (2013).

[44] E. Saloustros, G. Vichas, A. Margiolaki, Lung cancer in the era of Greek economic crisis, Lung Cancer 86 (2014) 1367–1380.

[45] B.H. Strand, A. Kunst, M. Huisman, G. Menvielle, M. Glickman, M. Bopp, et al., The reversed social gradient: higher breast cancer mortality in the higher educated compared to lower educated. A comparison of 11 European populations during the 1990s, Eur. J. Cancer 43 (7) (2007) 1200–1207.

[46] S. Gadeyne, G. Menvielle, I. Kulhanova, M. Bopp, P. Deboosere, T.A. Eikemo, et al., The turn of the gradient? Educational differences in breast cancer mortality in 18 European populations during the 2000s, Int. J. Cancer 141 (1) (2017) 33–44.

[47] J.A. McDougall, C.K. Blair, C.L. Wiggins, M.B. Goodwin, V.K. Chiu, A. Rajput, et al., Socioeconomic disparities in health-related quality of life among colorectal cancer survivors, J. Cancer Surviv. (2019).

[48] S.M. Cramb, K.L. Mengersen, G. Turrell, P.D. Baade, Spatial inequalities in colorectal and breast cancer survival: premature deaths and associated factors, Heal Place 18 (6) (2012) 1412–1421, https://doi.org/10.1016/j. healthplace.2012.07.006. Available from:.

[49] J.S. Haas, P. Brawarsky, A. Iyer, G.M. Fitzmaurice, B.A. Neville, C. Earle, Association of area sociodemographic characteristics and capacity for treatment with disparities in colorectal cancer care and mortality, Cancer (2011).

[50] M. Lian, M. Schootman, C.A. Doubeni, Y. Park, J.M. Major, R.A. Torres Stone, et al., Original Contribution Geographic Variation in Colorectal Cancer Survival and the Role of Small-Area Socioeconomic Deprivation: A Multilevel Survival Analysis of the NIH-AARP Diet and Health Study Cohort., Available from: https://academic. oup.com/aje/article-abstract/174/7/828/115994.

[51] P.D. Baade, P. Dasgupta, J.F. Aitken, G. Turrell, Geographic Remoteness, Area- level Socioeconomic Disadvantage and Inequalities in Colorectal Cancer Survival in Queensland: a Multilevel Analysis, Available from:, 2013 http://www.biomedce ntral.com/1471-2407/13/493.

[52] X. Li, K. Sundquist, J. Sundquist, Neighborhood deprivation and prostate cancer mortality: a multilevel analysis from Sweden, Prostate Cancer Prostatic. Dis. 15 (2011) 128–134. Available from: www.nature.com/pcan.

[53] Sanderson M., Coker A.L., Perez A., Du XL, Fadden MK, Health H. A Multilevel Analysis of Socioeconomic Status and Prostate Cancer Risk.

[54] K. Schwartz, I.J. Powell, W. Underwood Iii, J. George, C. Yee, M. Banerjee, Interplay of race, socioeconomic status and treatment on survival of prostate Cancer patients, Urology 74 (6) (2009) 1296–1302.

[55] Ministerio de Sanidad Servicios sociales e Igualdad, Los Sistemas Sanitarios En Los Países De La UE: Características E Indicadores De Salud 2013. Subdirección Gen Inf Sanit E Innovación, 99. Available from: https://www.mscbs.gob.es/estadEstudios/ estadisticas/docs/Sist_san.UE.XXI.pdf%0A, 2014, http://www.msssi.gob.es/estad Estudios/estadisticas/sisInfSanSNS/tablasEstadisticas/home.htm%5Cn2.

L. Palència et al.