SUU, Alm.del - 2015-16 - Endeligt svar på spørgsmål 819: MFU spm. om, hvilken forbindelse der er mellem danske patienters uddannelsesniveau og deres chancer for at være i live 1 år og 5 år efter, at kræftdiagnosen er stillet, til sundheds- og ældreministeren

Sundheds- og Ældreudvalget 2015-16
SUU Alm.del
Offentligt

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

available at www.sciencedirect.com

journal homepage: www.ejconline.com

Social inequality in incidence of and survival from cancer

in a population-based study in Denmark, 1994–2003:

Summary of ﬁndings

Susanne Oksbjerg Dalton

, Joachim Schuz

, Gerda Engholm

, Christoffer Johansen

Susanne Kruger Kjær

, Marianne Steding-Jessen

, Hans H. Storm

, Jørgen H. Olsen

Institute of Cancer Epidemiology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen Ø, Denmark

Department of Cancer Prevention and Documentation, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen Ø, Denmark

A R T I C L E I N F O

A B S T R A C T

Article history:

Received 6 June 2008

Received in revised form

10 June 2008

Accepted 16 June 2008

Available online 30 July 2008

Keywords:

Socioeconomic position

Denmark

Incidence

Survival

Cancer

The purpose of this nationwide, population register-based study was to describe variations

in cancer incidence and survival by social position in a social welfare state, Denmark, on

the basis of a range of socioeconomic, demographic and health-related indicators. Our

study population comprised all 3.22 million Danish residents born in 1925–1973 and aged

P30

years, who were followed up for cancer incidence in 1994–2003 and for survival in

1994–2006, yielding 147,973 cancers. The incidence increased with lower education and

income, especially for tobacco- and other lifestyle-related cancers, although for cancers

of the breast and prostate and malignant melanoma the association was inverse. Con-

versely there was a general increase in incidence among early retirement pensioners, per-

sons living in rented housing and those living in the smallest dwellings. Also incidence

rates were generally higher in persons living alone compared to those living with a partner

and in the capital area compared to the rural areas. Social inequality in the prognosis of

most cancers was observed, despite the equal access to health care in Denmark, with

poorer relative survival related to fewer advantages, regardless of how they were measured,

often most pronounced in the ﬁrst year after diagnosis. Also living alone and having

somatic or psychiatric comorbidity negatively impacted the relative survival after most

cancers. Our study shows that inequalities in cancer incidence and survival must be

addressed in all aspects of public health, with interventions both to reduce incidence

and to prolong survival.

Background

Socioeconomic position is a commonly used concept in

health research, and general inequalities in health reﬂect so-

cial inequalities. An extensive review from the late 1990s pro-

vided clear evidence that both cancer incidence and survival

are related to social position, lower social position tending to

be associated with higher cancer incidence and poorer cancer

survival, even if the pattern differs by cancer site.

Although

Denmark has an extensive welfare system and equal access

to health care, social inequalities in both incidence of and

survival from some cancers have been observed.

2–4

Socioeconomic position is based on the social and eco-

nomic factors that inﬂuence the positions individuals or

Corresponding author:

Tel.: +45 35257500; fax: +45 35257731.

E-mail address:

[email protected]

(S.O. Dalton).

0959-8049/$ - see front matter

doi:10.1016/j.ejca.2008.06.018

SUU, Alm.del - 2015-16 - Endeligt svar på spørgsmål 819: MFU spm. om, hvilken forbindelse der er mellem danske patienters uddannelsesniveau og deres chancer for at være i live 1 år og 5 år efter, at kræftdiagnosen er stillet, til sundheds- og ældreministeren

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

2075

groups hold in the structure of a society. Each indicator of

socioeconomic position commonly used, such as education,

income and occupation, measures a different but related as-

pect and may be more or less relevant to the outcome being

studied. In addition to social gradients by education and in-

come, there may be particularly vulnerable groups, who can

be identiﬁed by other social or health-related indicators, such

as early retirement pensioners, persons living in small apart-

ments or peripheral rural areas or persons with comorbid

conditions.

In this nationwide descriptive study, we systematically

examined variations in the incidence of and survival from

cancer in Denmark on the basis of a range of socioeconomic,

demographic and health-related indicators, with the aim of

addressing the following questions:What is the effect of indi-

cators of social position on the incidence of cancer? To what

extent do these socioeconomic indicators affect survival after

cancer? Are some indicators of social position associated with

a low incidence of cancer and good chances for survival after

cancers at speciﬁc sites? Are there, conversely, indicators of

social position that are markers for a high incidence of cancer

and a worse prognosis?

The purpose of this paper is to summarise the ﬁndings

from the analyses of cancers at speciﬁc sites, highlight pat-

terns of social inequality in incidence and survival and direct

attention to groups who bear a disproportionately heavy bur-

den of cancer in Danish society today. The results of this

study are important for other industrialised societies, because

they mirror a society characterised by a high level of public

social security. In general, Danes, although a small popula-

tion, have organised a society that in many respects is seen

as a model of the modern social welfare state.

range of the cohort, represents 58% of all cancers diagnosed

in Denmark during that period. For each level of each indica-

tor, we conducted relative survival analyses, adjusting for le-

vel-speciﬁc population mortality amongst the incident cancer

cases in 1994–2003 with follow-up through 2006.

Population

mortality rates were stratiﬁed by age, period and the respec-

tive indicator. Except for the analyses of ethnicity, all analyses

included only residents born in Denmark to at least one Dan-

ish-born parent with Danish citizenship.

In this paper, we present selected estimated IRRs for the

cancer sites included and for all cancers in relation to levels

of the socioeconomic, demographic and health-related indi-

cators for 1-year and 5-year relative survival. We thus provide

an overview of the ﬁndings for both incidence and survival for

the cancers at the sites investigated and evaluate patterns of

social inequality and indications of groups who are vulnerable

to cancer or to a worse prognosis. More detailed information

on the IRRs and relative survival estimates for each cancer

site is provided in separate papers in this supplement to

The European Journal of Cancer.

7–17

Social inequality in the incidence of cancer

Material and methods

The material and methods are described elsewhere.

Brieﬂy,

the study population comprised all 3.22 million Danish resi-

dents born between 1925 and 1973, and who entered the co-

hort at age 30 without a previous cancer (see Fig. 1 in

[5])

Information on socioeconomic, demographic and health-re-

lated indicators was obtained from various Danish registers

based on administrative data. The indicators were level of edu-

cation, disposable income, afﬁliation to the work market, so-

cial class, housing tenure, size of dwelling, cohabitation

status, type of district, ethnicity, Charlson comorbidity index,

depression and schizophrenia measured at the individual le-

vel on an annual basis (see Table 1 in

[5]

for deﬁnition of the

variables). Crude, age-speciﬁc and age-standardised incidence

rates were calculated for 21 cancers at selected sites (see Table

4 in

[5])

and for all cancers combined, diagnosed in the cohort

in 1994–2003. The incidence rates were standardised by age (in

5-year age groups) and period (in two 5-year periods), with the

total study population as the standard.

Further, we used log-

linear Poisson regression to model incidence rate ratios (IRRs),

ﬁrst adjusted for period (in 5-year periods) and age (as two con-

tinuous variables: age and age

in years) and secondly by add-

ing education and disposable income to the models.

The study is based on 3,218,440 persons and 25,764,811

person–years of follow-up for cancer. A total of 147,973 can-

cers were diagnosed in the cohort, which, due to the age

Table 1

provides a summary of the associations between edu-

cational level and disposable income and the incidence rates

of cancer, by site. We show the estimated IRRs for persons with

basic or high-school education as compared with higher edu-

cation and with low income (1st quartile) as compared with

middle income (2nd and 3rd quartiles), each adjusted for each

other. A consistent increase in the incidence rates was ob-

served with shorter education and lower income for cancers

of the mouth and pharynx, larynx, oesophagus, stomach, lung,

kidney and cervix, all of which are related to tobacco smoking,

whereas inverse associations were found for breast cancer,

prostate cancer and malignant melanoma, with increasing

incidences with longer education and higher income. For pan-

creas and bladder cancer, associations were found with educa-

tion but not with income. For the remaining cancers sites, such

as colon and rectum cancer, brain or central nervous system

cancer and haematological cancers in both sexes and with cor-

pus and ovary cancer in women and testicular cancer in men,

no clear associations were found between incidence and socio-

economic position. The incidence of all cancers showed a

small but statistically signiﬁcant excess amongst men, but

not women, with low social position (Table

1).

Table 2

provides an overview of the associations between

indicators that are markers for possibly disadvantaged or vul-

nerable groups in regard to cancer incidence. A general pat-

tern in regard to cancer incidence was seen for early

retirement pensioners, persons living in rented housing and

those living in the smallest dwellings. The ﬁrst group had a

particularly high incidence rate of lung cancer. Further, high-

er IRR estimates were observed for divorced persons than

those who were married, and a similar pattern was seen for

single persons and widows and widowers (data not shown).

An urban–rural gradient in the incidence rates of most can-

cers was found, with the highest rates in the capital area

and lower rates amongst persons living in peripheral rural

areas (Table

2).

As the group of immigrants and their descen-

dants was small and heterogeneous, we did not include them

2076

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

Table 1 – Incidence rate ratios by cancer site for indicators of low social position, basic school or high-school education and

low income in persons born in 1925–1973 and aged

P30

years in Denmark between 1994 and 2003

Cancer site

Basic or high-school

education (adjusted IRR)

Men

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

1.43

1.67

1.30

1.37

1.20

0.93

1.02

1.53

Low income

(adjusted IRR)

Men

1.74*

1.23

1.16

1.20

0.98

0.99

1.04

1.21

Overall association

with low social status

Women

1.25

3.23

0.87

1.23

1.22

1.02

1.12

1.85

0.80

1.33

0.98

0.97

Women

1.25

1.18

1.14

1.03

0.94

0.95

1.04

1.06

0.95

1.13

0.94

0.98

"/!

0.81

1.00

1.22

1.15

0.65

1.04

1.10

1.05

0.96

1.10

1.54

1.37

0.69

0.92

1.14

1.16

1.10

1.02

0.92

0.91

1.13

0.97

0.88

1.02

1.01

1.24

1.04

1.07

1.10

0.95

0.96

1.05

0.98

0.92

1.01

1.00

Adjusted IRR, incidence rate ratios adjusted for age, period, level of education and disposable income; basic or high-school education compared

with higher education, and low income compared with middle income.

95% conﬁdence interval excludes 1.

indicates increased incidence rate ratio with level of indicator in question;

indicates reduced incidence rate ratio with level of indicator in

question;

indicates no change in incidence rate ratio with level of indicator in question; / indicates that education and income have different

patterns.

in the tables; however, those from non-western countries ap-

peared to have lower incidence rates for most cancers than

the total group of Danes, except for stomach cancer,

for

which the incidence rates were increased in the former group.

Health-related indicators, such as the Charlson comorbid-

ity index, depression and schizophrenia, were included in

the study in order to obtain estimates of cancer incidence in

persons with major health problems (Table

3).

Having chronic

somatic disorders, as measured by the Charlson comorbidity

index, resulted in increased incidence rates for smoking-re-

lated cancers (cancers of the mouth and pharynx, larynx, lung,

oesophagus, stomach, pancreas, kidney and bladder) and for

haematological cancers (Table

3).

Having been hospitalised

for depression was associated with increased incidence rates

for some smoking-related cancers, i.e. cancers of the mouth

and pharynx, larynx and lung in men and women and oesoph-

agus cancer in women. The incidence rates of breast cancer

and brain cancer were also increased in women with depres-

sion. In persons with schizophrenia, the incidence rates of

lung, breast and cervix cancer were increased and the inci-

dence rate of prostate cancer was reduced (Table

3).

Social inequality in survival after cancer

Tables 4–7

show the estimates of relative survival (in %) for

the levels of each indicator corresponding to the lowest and

highest socioeconomic positions, for example, we compared

the relative survival of persons with basic or high-school edu-

cation with that of persons with higher education.

Table 4

shows that for most cancers there were marked differences

in relative survival by the level of education and disposable

income, both 1 and 5 years after cancer diagnosis, persons

with the least education and the lowest income having poorer

relative survival. Sometimes persons with vocational educa-

tion had the best survival, after 1 year for lung cancer,

after

5 years for colon cancer amongst women,

and after 1 year for

non-Hodgkin lymphoma.

For all cancers together, an approximately 10% difference

was already apparent after 1 year; the difference was only

1–2% greater after 5 years. This overall 10% difference was

greater than for most individual cancer sites, as it is a com-

bined effect of more cancers with an overall poor prognosis

and poorer relative survival in persons with low social

position.

For some cancers, including those at sites with poorer sur-

vival, such as cancers of the oesophagus and lung, but also

cancers with better prognosis, such as cancers of the ovary

and rectum and leukaemia, most of the differences in relative

survival between social groups were observed after 1 year,

with less difference or none in long-term relative survival.

Early retirement pensioners were identiﬁed as having

poorer relative survival in both the short and the long term

Table 2 – Incidence rate ratios by cancer site for indicators of disadvantaged groups in persons born in 1925–1973 and aged

P30

years in Denmark between 1994 and 2003,

by sex and cancer site

Cancer site

Early retirement

pensioners (adjusted IRR)

Men

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

Overall pattern

4.52

2.75

2.43

1.28

1.65

1.18

8.56

Rented housing

(adjusted IRR)

Men

2.83

2.27

1.90

1.19

1.40

1.19

1.17

1.67

Dwelling < 50 m

(adjusted IRR)

Men

4.08

2.57

2.26

1.28

1.51

1.22

1.10

2.12

Divorced

(adjusted IRR)

Men

3.13

2.41

1.95

1.05

1.35

1.14

1.72

Peripheral rural area

(adjusted IRR)

Men

0.52

0.62

0.59

0.87

0.82

0.90

0.83

1.05

Women

3.50

2.15

2.62

1.43

1.34

1.02

1.08

4.52

1.03

1.31

0.93

0.95

Women

2.19

2.02

1.54

1.21

1.33

0.98

1.04

1.57

1.04

1.51

0.94

1.05

Women

2.76

0.81

1.97

0.92

1.08

0.91

0.97

2.95

1.02

1.55

0.83

0.98

Women

2.14

1.96

1.39

1.56

1.35

0.94

1.79

1.04

1.90

0.78

0.99

Women

EUROPEAN JOURNAL OF CANCER

0.48

0.78

0.53

0.77

0.80

0.92

1.03

1.02

0.89

0.75

0.97

1.04

0.87

1.10

1.16

1.24

0.60

1.05

1.14

1.26

1.17

5.28

1.29

1.36

0.67

1.10

1.31

1.10

0.99

2.57

0.98

1.10

1.20

1.21

0.88

0.99

1.06

1.29

0.97

1.31

1.17

1.34

0.85

1.05

1.11

1.00

0.93

1.14

0.88

0.80

1.19

1.16

0.68

0.72

0.96

1.06

1.26

1.53

0.81

1.00

0.75

0.76

0.86

1.41

0.86

1.37

!/"?

0.97

1.04

1.24

1.01

0.82

0.91

1.03

0.89

0.93

1.32

0.96

1.34

0.84

0.96

1.04

0.96

0.98

1.16

0.86

0.89

0.91

0.83

0.95

0.85

0.91

0.80

0.83

0.87

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

1.03

0.76

1.03

0.87

0.90

1.07

0.95

0.93

Adjusted IRR, incidence rate ratios adjusted for age, period, level of education and disposable income; early retirement pensioners compared with persons at work; rented housing compared with

owner-occupied housing; dwelling < 50 m

compared with dwelling 100–149 m

; divorced compared with married; peripheral rural area compared with capital area.

95% conﬁdence interval excludes 1.

indicates increased incidence rate ratio with level of indicator in question;

indicates reduced incidence rate ratio with level of indicator in question;

indicates no change in incidence rate ratio with

level of indicator in question.

2077

2078

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

Table 3 – Incidence rate ratios by cancer site for health-related indicators in persons born in 1925–1973 and aged

P30

years

in Denmark between 1994 and 2003, by sex

Cancer site

Charlson

2 (adjusted IRR)

Men

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

Overall pattern

2.17

1.74

1.53

1.35

1.44

1.24

1.04

2.01

Depression (adjusted IRR)

Men

1.64

1.32

1.09

1.16

0.84

1.01

0.90

2.15

Schizophrenia (adjusted IRR)

Men

1.09

1.01

1.50

1.32

0.88

0.95

0.70

3.03

Women

2.75

2.76

2.03

1.54

1.55

1.10

1.00

2.69

1.01

1.08

0.82

0.94

Women

1.55

1.50

2.06

1.03

1.27

0.92

0.97

2.01

1.10

1.00

0.71

0.94

Women

0.87

0.73

1.41

1.32

0.88

0.92

1.04

2.51

1.18

1.35

0.87

0.98

1.03

1.27

1.57

1.44

0.92

1.20

1.63

2.16

1.24

1.47

1.54

1.63

0.87

1.13

1.94

2.41

1.14

1.39

0.93

0.78

0.91

1.09

1.01

1.10

0.99

1.10

1.29

(")

1.25

1.21

0.95

1.27

0.75

1.08

0.77

1.21

(")

0.59

0.80

1.35

0.87

0.86

0.78

1.21

0.96

1.05

1.47

(")

1.31

0.83

0.81

0.89

1.13

1.69

1.19

1.40

(")

Adjusted IRR, incidence rate ratios adjusted for age, period, level of education and disposable income; Charlson

2 compared with Charlson 0;

depression compared with no depression; schizophrenia compared with no schizophrenia.

95% conﬁdence interval excludes 1.

indicates increased incidence rate ratio with level of indicator in question;

indicates reduced incidence rate ratio with level of indicator in

question;

indicates no change in incidence rate ratio with level of indicator in question; (") indicates incidence rate ratios not increased for

most sites but overall increase due to increases for major sites.

(Table

5).

For all cancers together, the difference in relative

survival between early retirement pensioners and persons ac-

tive in the work market was almost 20%, again with a com-

bined effect of more cancers with a poor prognosis and

poorer relative survival from cancers at most sites. Relative

survival was also poorer amongst persons living in rented

rather than owned housing and for those who lived in smaller

rather than larger housing (Table

5).

With respect to demographic indicators for short-term

survival, being married was generally associated with longer

relative survival, although most of the conﬁdence intervals

overlapped (Table

6).

The difference was more pronounced

for men than for women with colon and rectum cancer and

some of the smoking-related cancers, such as of the mouth

and pharynx, larynx, kidney and urinary bladder. Divorced

persons (Table

6),

widows and widowers and single persons

had poorer relative survival than married people from can-

cers at many sites (data not shown). No general pattern in rel-

ative survival from any cancer could be discerned by type of

district (Table

6).

The relative survival of immigrants or their

descendants from non-western countries seemed to be, if

anything, better than that of the combined group of Danes;

however, the estimates for these heterogeneous groups must

be interpreted cautiously (data not shown).

Having somatic or psychiatric comorbidity appeared to be

associated with poorer relative survival than for patients

without, but the conﬁdence intervals overlapped for most

sites (Table

7).

For cancers such as those of the larynx and uri-

nary bladder cancer and malignant melanoma and non-

Hodgkin lymphoma, men with a score of 2 or more on the

Charlson comorbidity index had signiﬁcantly poorer 5-year

relative survival than men with a score of 0 (Table

7).

The rel-

ative survival of cancer patients with depression or schizo-

phrenia was signiﬁcantly worse for only a few sites, i.e. for

men with stomach cancer (1-year relative survival), for breast,

ovary and rectum cancer in women (1-year relative survival)

and for cervical cancer both 1 and 5 years after diagnosis.

Comment

In this paper, we have summarised the ﬁndings of a nation-

wide population-based study of the effects of socioeconomic,

demographic and health-related indicators and on the inci-

dence of and survival from cancers diagnosed between 1994

and 2003 in Denmark. Access to complete nationwide data

sources based on data collected for administrative purposes

and covering the entire Danish population and use of the un-

ique personal identiﬁcation numbers ensured a design that is

virtually free of information or selection bias and essentially

eliminates loss to follow-up. The outcome studied was ob-

tained from the oldest cancer register in the world, estab-

lished in 1943.

We were able to obtain individual

information on social and economic status, demography

and health at a level of detail that is unique, as access to

Table 4 – Indicators of social position and relative 1-year and 5-year survival (in %), by cancer site, in persons born in 1925–1973 and aged

P30

years in Denmark between

1994 and 2006

Cancer site

Men

1-year RS

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

66/78

85/89

27/36

39/41

14/16

69/75

77/84

28/34

Level of education

Basic or high-school/higher

Women

5-year RS

30/39

52/59

5/7

13/10

2/3

42/46

44/50

7/10

Disposable income

Low/high

EUROPEAN JOURNAL OF CANCER

Men

1-year RS

66/80

81/94

25/34

37/41

17/20

67/77

75/84

27/33

Women

5-year RS

25/46

45/63

4/7

12/13

3/3

40/46

41/51

7/8

1-year RS

76/73

81/76

30/52

38/38

16/15

73/77

81/88

33/34

95/98

88/94

93/95

74/81

5-year RS

44/47

54/65

6/16

16/15

2/2

46/49

51/57

9/10

77/84

68/78

79/81

37/36

1-year RS

68/73

77/90

31/29

37/42

15/16

73/79

82/87

32/36

95/97

87/92

92/96

73/82

5-year RS

42/43

50/71

5/8

15/18

1/3

45/55

49/58

9/10

75/83

68/73

77/83

36/39

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

88/92

98/99

62/68

86/93

93/97

60/69

73/79

90/93

72/80

62/73

47/59

93/97

38/41

68/75

75/81

39/47

48/58

82/76

46/54

37/48

61/74

81/89

97/98

72/79

80/85

90/97

71/78

72/82

42/49

62/70

86/92

58/66

58/65

78/90

46/52

50/62

88/93

97/99

62/68

85/93

91/96

60/67

72/79

91/91

72/78

60/73

47/56

93/97

37/44

65/74

73/82

42/43

46/56

78/78

45/56

34/48

60/70

81/86

97/98

72/76

83/83

93/94

73/79

72/82

38/50

60/66

87/92

58/65

59/68

81/87

49/57

49/62

RS, cumulative level-speciﬁc relative survival in %.

Basic or high-school compared with higher education, and low income compared with high income.

95% conﬁdence intervals non-overlapping between levels of indicator.

2079

2080

Table 5 – Indicators of social position and relative 1-year and 5-year survival (in %), by cancer site, in persons born in 1925–1973 and aged

P30

years in Denmark between

1994 and 2006

Cancer site

Afﬁliation to work market

Early retirement pensioner/working

Men

1-year

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

64/82

80/90

24/36

35/44

19/11

60/77

79/84

23/37

Housing tenure

Rental/owner-occupied

Men

1-year

64/78

81/90

30/30

39/40

12/18

66/75

77/80

28/31

Dwelling size (m

)

0–49/> 150

Men

1-year

56/78

79/93

19/33

38/42

14/19

62/77

71/82

18/33

Women

1-year

63/85

70/91

26/43

34/43

16/22

66/80

84/88

35/41

94/98

78/94

93/96

71/88

Women

5-year

30/38

48/60

6/7

14/12

2/3

39/46

43/48

7/8

Women

5-year

21/42

38/65

3/5

16/15

7/2

36/49

41/48

4/9

EUROPEAN JOURNAL OF CANCER

5-year

27/46

40/59

4/8

12/15

3/3

35/48

44/51

8/10

5-year

35/55

38/70

4/13

15/20

2/4

39/52

51/57

10/11

76/83

55/79

86/87

35/48

1-year

72/76

82/79

31/33

34/39

16/16

72/76

81/83

33/34

95/97

87/91

92/94

76/78

5-year

41/45

54/57

6/9

10/18

2/2

47/49

50/55

9/9

77/80

68/73

78/81

38/46

1-year

81/77

102/90

32/35

17/42

12/16

53/78

78/86

25/37

94/97

69/91

93/96

64/81

5-year

29/50

55/68

0/9

9/20

9/3

37/51

29/59

6/11

73/81

56/73

82/83

34/39

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

91/91

94/99

56/72

88/93

97/97

59/73

68/85

82/95

72/81

53/74

44/50

84/96

35/49

69/81

78/84

39/50

43/64

86/87

44/58

31/52

57/79

80/93

95/98

75/83

75/90

72/98

64/81

70/87

49/58

65/80

93/92

64/69

57/76

64/91

43/58

50/68

87/91

97/98

57/67

85/90

94/95

59/64

72/79

87/92

71/76

59/68

51/52

92/96

37/41

66/70

77/80

39/42

45/56

74/81

44/51

35/43

64/66

79/85

97/98

74/75

82/82

89/95

69/74

72/77

43/44

60/67

88/89

60/62

58/62

77/85

45/49

50/56

80/92

98/97

55/68

75/91

91/95

51/66

67/81

94/92

69/76

52/61

38/55

85/94

40/42

53/71

67/81

39/45

48/57

89/79

34/52

29/46

83/66

92/86

99/97

62/76

57/86

100/97

85/75

65/80

76/43

65/70

94/90

56/63

49/66

101/89

69/47

45/59

RS, cumulative level-speciﬁc relative survival in %.

Early retirement pensioners compared with being at work, rented housing compared with owner-occupied housing and small dwellings (0–49 m

) compared with large dwellings (Pm

95% conﬁdence intervals non-overlapping between levels of indicator.

Table 6 – Demographic indicators of social position and relative 1-year and 5-year survival (in %), by cancer site, in persons born in 1925 and 1973 and aged

P30

years in

Denmark between 1994 and 2006

Cancer site

Men

1-year RS

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

64/76

77/90

23/32

33/41

13/16

67/74

73/81

24/32

Cohabitation status

Divorced/married

Women

5-year RS

27/40

37/61

4/7

9/14

2/3

40/46

35/49

6/8

Type of district

Peripheral rural area/capital area

Men

1-year RS

69/79

85/83

17/30

37/40

21/14

73/72

81/80

32/29

Women

5-year RS

29/31

55/50

3/5

12/12

5/3

41/43

50/46

8/7

EUROPEAN JOURNAL OF CANCER

1-year RS

69/78

78/82

23/36

31/39

15/16

75/75

79/83

34/34

96/97

86/91

93/94

74/80

5-year RS

39/46

41/56

2/11

11/17

1/3

46/49

48/55

8/9

78/80

66/72

79/81

32/38

1-year RS

86/69

75/77

46/33

48/35

12/14

72/75

80/81

35/33

97/96

87/90

95/94

77/79

5-year RS

61/40

49/53

10/5

24/12

0/2

50/50

48/52

11/9

79/80

68/71

83/80

41/35

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

87/91

97/98

52/67

84/90

92/95

57/63

71/81

81/92

72/75

56/68

48/53

91/95

28/42

63/70

77/80

38/41

49/56

64/79

49/52

31/43

62/68

82/84

96/98

70/74

86/84

88/96

70/73

71/77

42/46

62/66

89/90

54/61

60/63

88/86

48/49

48/56

88/91

99/98

60/61

87/87

93/96

63/61

76/77

75/83

64/76

64/65

43/57

99/95

37/39

66/69

81/82

40/39

56/51

58/71

34/52

38/41

66/68

81/84

100/97

70/75

78/82

100/94

66/77

75/76

50/49

63/65

92/89

57/61

48/62

68/85

43/55

54/55

RS, cumulative level-speciﬁc relative survival in %.

Divorced compared with married and peripheral rural areas compared with capital area.

95% conﬁdence intervals non-overlapping between levels of indicator.

2081

2082

Table 7 – Indicators of health and relative 1-year and 5-year survival (in %), by cancer site, in persons born in 1925–1973 and aged

P30

years in Denmark between 1994

and 2006

Cancer site

Charlson comorbidity

index

P2/0

Men

1-year

Mouth and pharynx

Larynx

Oesophagus

Stomach

Pancreas

Colon

Rectum

Lung

Breast

Cervix

Corpus

Ovary

Prostate

Testis

Kidney

Bladder

Malignant melanoma

Brain and central nervous system

Non-Hodgkin lymphoma

Hodgkin lymphoma

Leukaemias

All cancers

58/73

73/88

25/31

32/40

11/16

66/73

75/80

27/30

Depression

Yes/no

Schizophrenia or

other psychosis

Yes/no

Women

Men

1-year

66/70

95/85

31/30

32/40

0/16

58/72

51/79

21/30

Women

1-year

64/78

75/79

22/34

35/36

8/16

70/75

78/83

31/34

95/96

81/90

90/94

63/79

Men

1-year

70/70

86/85

17/30

16/40

8/16

67/72

79/79

29/30

Women

EUROPEAN JOURNAL OF CANCER

5-year

27/35

38/56

7/6

7/13

2/3

41/45

42/47

7/8

5-year

29/46

61/58

5/8

9/15

2/2

46/48

51/52

7/9

76/79

58/72

76/79

32/38

5-year

31/33

52/54

5/7

8/13

0/3

40/44

45/46

6/8

1-year

72/74

77/81

38/32

17/38

17/16

77/74

86/82

34/34

96/96

80/89

93/94

76/77

5-year

37/42

57/56

13/7

0/15

2/2

50/48

51/53

10/9

78/79

60/71

80/80

33/37

5-year

7/33

77/53

0/7

0/13

0/3

28/44

27/46

5/8

1-year

64/74

104/80

19/33

8/38

14/16

76/74

60/83

29/34

93/96

66/90

81/94

58/77

5-year

69/42

121/55

15/8

0/15

0/2

47/48

34/53

6/9

74/79

46/71

64/80

22/37

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

89/90

95/98

64/65

83/89

85/95

62/63

57/79

83/90

61/76

56/67

51/52

94/95

34/40

59/71

58/79

38/42

39/54

83/78

34/51

32/42

57/66

73/84

96/98

73/75

75/84

79/96

69/74

64/77

40/43

46/66

88/89

55/62

51/63

81/85

38/49

43/56

88/90

89/98

61/64

85/88

97/95

70/62

87/77

101/90

80/74

60/65

54/52

92/95

35/39

70/69

66/79

53/41

57/53

58/78

49/49

35/40

61/65

75/83

97/98

74/74

83/82

77/93

53/73

70/75

41/44

64/64

85/89

63/61

61/61

60/83

41/48

48/54

80/90

101/98

44/64

80/88

100/95

48/62

66/77

101/90

45/74

48/65

40/52

98/95

38/39

68/69

84/79

28/41

39/53

107/78

36/49

28/40

69/65

75/83

98/98

76/74

74/82

100/93

46/73

66/75

58/44

63/64

70/89

63/61

57/61

102/82

38/48

44/54

RS, cumulative level-speciﬁc relative survival in %.

Charlson comorbidity index score

compared with score 0, depression compared with no depression and schizophrenia or other psychosis compared with no schizophrenia or other psychosis.

*95% conﬁdence intervals non-overlapping between levels of indicator.

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

2083

national statistics for scientiﬁc purposes is possible in only a

few countries in the world.

The age limits for the cohort were deﬁned to cover persons

of working age and those who had recently left the labour

market. This was partly because the reliability and availability

of data on socioeconomic status in Danish registers are

poorer for the oldest section of the population. As we aimed

to include the most recent population data, we analysed can-

cer incidence over a rather short period, some 10 years.

Although this limitation precludes analyses of trends over

time, we obtained an updated ‘snapshot’ of the importance

of different indicators of social position for the incidence of

and survival from cancer in Danish society today.

We saw relatively clear patterns of social inequality in the

incidence rates of especially the smoking-related cancers

(cancers of the mouth and pharynx, larynx, oesophagus, pan-

creas, lung, kidney, cervix and urinary bladder).

There was

thus a stepwise increase in the incidence rates with decreas-

ing social position, measured as both level of education and

disposable income. Even after adjustment for these two fac-

tors, however, an effect was still seen for poorer afﬁliation

to the labour market and living in small housing, as shown

in more detail in the site-speciﬁc papers. Early retirement

pensioners in particular had higher incidence rates of many

cancers, particularly lung cancer. Although some 30% of the

person–years were accumulated by persons living in rented

housing, this factor still resulted in increased incidence rates

for many cancer sites after adjustment for education and in-

come, suggesting that housing tenure, which is a measure of

material aspects of socioeconomic circumstances,

is associ-

ated with aspects of social position other than education and

income. For breast cancer, prostate cancer and malignant

melanoma, the situation was generally reversed, with

increasing incidence rates associated with higher social posi-

tion.

11,13,15

In addition, there was a slight but signiﬁcantly

higher (5%) incidence rate of breast cancer amongst women

receiving early retirement pension, which might partly reﬂect

the higher IRR for breast cancer observed in women with

either depression or schizophrenia.

The incidence rate differences calculated by level of edu-

cation provide an overall estimate of the potential for preven-

tion, which is large for many cancer sites at a population

level. We estimated that the incidence rates of most smok-

ing-related cancers could be reduced by 25–50% in the group

with basic or high-school education if they have the same risk

factor proﬁle as persons with higher education, and this ap-

plies to both men and women.

7,8,10,14

Likewise, the incidence

rates of breast cancer, prostate cancer and malignant mela-

noma in the group with higher education could be reduced

by 20–40% if they have the same risk factor proﬁle as the

group with basic or high-school education.

11,13,15

Many of the risk factors for cancer, such as smoking, alco-

hol intake, unhealthy diet and a sedentary lifestyle, are con-

centrated in less advantaged groups.

21–23

It is therefore not

surprising that most of the cancers for which relatively large

incidence rate differences were seen were cancers with a

strong lifestyle component, such as smoking-related cancers.

The magnitude of the differences observed not only indicates

a large potential for prevention but also poses challenges, as

some of the lifestyle factors that probably play a role in the

social gradients observed in this study might be distributed

even more unevenly amongst social groups in the future, as

lifestyle is closely linked to the material and cultural re-

sources that make up socioeconomic position. Nevertheless,

prevention initiatives targeted at socially disadvantaged

groups might have a strong effect on public health. Cancers

that occur at higher incidence amongst afﬂuent persons than

those with lower social status represent another area for pre-

ventive policies. There is currently no scientiﬁc evidence to

explain why prostate cancer occurs in excess amongst men

of high social status. Most studies have identiﬁed reproduc-

tive factors as the explanation for the high incidence of breast

cancer in afﬂuent women. Likewise, ﬁnancial ease is probably

the explanation for the fact that persons of high socioeco-

nomic status travel more, which might explain their higher

incidence of melanoma. There is clear preventive potential

for this cancer.

The pattern of social inequality in relative survival from

cancer is more uniform across sites with survival being re-

duced in disadvantaged groups or no differences amongst

groups. For some cancers, mainly those with poorer sur-

vival such as of the oesophagus and lung but also cancers

of the ovary and kidney, differences amongst social groups

were observed only for the short term. For other cancers,

the inequalities remained throughout the 5 years of obser-

vation, including cancers of the colon, breast, cervix, pros-

tate and urinary bladder, and malignant melanoma and

non-Hodgkin lymphoma. Information on tumour stage

was not available; however, cancer patients of higher socio-

economic position might have beneﬁtted from, e.g. earlier

diagnosis, which would result in differential stage distribu-

tion by socioeconomic position and might explain some of

the differences in short-term survival from some cancers.

Other explanations might be differences in peri-operative

mortality and in access to state-of-art treatment,

24–27

by so-

cial position. Some socioeconomic factors, such as afﬁlia-

tion to the work market and educational level, did in fact

predict a delay in cancer diagnosis in a Danish popula-

tion-based cohort study of cancer patients and their general

practitioners.

This is similar to the ﬁndings from a UK

population-based study that showed that delay of diagnosis

and treatment was longer in lower socioeconomic groups

than in higher socioeconomic groups.

Comorbidity was a marker for worse relative survival from

cancers at many sites. In the previous studies also, comorbid-

ity had been found to be related to stage of and survival from

cancers of the breast, ovary and prostate in Denmark.

3,30–32

Presence of concurrent disease at the time of cancer diagnosis

might lead to incomplete staging and less tolerance for the

necessary cancer treatment, and thus possibly inﬂuence ther-

apeutic decision making. We were unable to examine

whether comorbidity had differential prognostic impact by

socioeconomic position, but our ﬁnding that comorbidity

inﬂuenced survival after most cancers is compatible with

other population-based studies, i.e. based on the Eindhoven

Cancer Registry where an independent prognostic effect of

comorbidity for several cancers was observed.

33–35

Improved

access to or compliance with treatment of comorbid disorders

must be a priority in cancer patients of all socioeconomic

groups.

2084

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

The number of cancer cases for which there is social

inequality in long-term survival is relatively large. Cancers

of the colon, breast and prostate, and malignant melanoma

and non-Hodgkin lymphoma constitute roughly one-third of

all cancer cases in Denmark, and interventions must there-

fore target social inequality in the management of these can-

cers. We had no information on access to or compliance with

treatment regimens or information from clinical monitoring

of outpatients. Therefore, the observed differences in relative

survival might be due to differences in access to treatment,

including second-line or experimental treatment, and to dif-

ferences in stage distribution, despite the supposedly equal

access to treatment and care in Denmark.

Further, it is likely that social inequalities in health in gen-

eral,

36,37

play a role in the long-term differences in survival

from cancer. In the Eurocare studies, cancer patients in Den-

mark consistently had the poorest survival of all the countries

in northern Europe, the survival estimates being closer to the

average of countries in western Europe and close to estimates

in England and Scotland, although better than those for east-

ern Europe.

38–40

For cancers of the stomach, colon, rectum,

lung, breast, prostate and all sites combined, the Danish 5-

year relative survival rate is below the European average; only

the rates for survival from cancers of the cervix and testis and

malignant melanoma are above the European average.

These ﬁndings, especially in comparison with other northern

European countries, contributed importantly to the formula-

tion of the national action plan against cancer. Two of the

strongest measures of social position, level of education and

disposable income, indicate clearly that survival from cancers

of the lung, colon, rectum, breast, prostate and cervix

amongst groups with higher social status is equal to or even

better than the survival observed in the other countries of

northern Europe.

The high incidence of and poor survival

from smoking-related cancers show clearly that the very high

prevalence of smoking in Denmark over the past several dec-

ades has played an important role.

Tobacco use is, however, not the only explanation, as Dan-

ish patients with cancers not related to smoking (breast, mel-

anoma, prostate) also had poorer survival compared to those

in northern Europe, although socially advantaged groups had

better survival.

11,13,15

For cancer of the prostate, the survival

of persons of high social status was similar to that seen in

the other Nordic countries, where testing for prostate-speciﬁc

antigen (PSA) is common and the incidence rates are high,

which may indicate inequality in the use of health care. There

is no formal screening programme for prostate cancer in Den-

mark, and the extent of PSA screening amongst Danish men

has been limited.

41,42

It is likely that men of higher social sta-

tus are better informed and, consequently, more aware of the

symptoms and the possibility of asymptomatic (latent) pros-

tate cancer. They would therefore be more likely to ask for a

PSA test. As a result, differential ‘grey zone’ screening might

inﬂuence not only the incidence but also the survival pattern

by social group.

Our ﬁndings for relative survival from cervix cancer are

interesting in this context. In the Eurocare project, the rates

in Denmark are better than the average

; however, when

only Danish patients of higher social position are consid-

ered,

the survival estimates exceed even those of Iceland,

which has the best survival rates in Europe.

Compliance

with the screening programme for cervix cancer in Denmark

is fairly good but is likely to be inﬂuenced by social status, as

reﬂected in both short- and long-term inequalities. Well-

organised programmes for early detection with high compli-

ance might limit social differences in cancer care. Thus, the

challenge is to increase participation in or procedures for

early detection and to improve treatment for all social groups.

This study covers an entire nation, which for historical

and cultural reasons has striven to ensure equality in all as-

pects of life. This political and administrative goal covers all

facets of public social support and preventive public health

interventions; however, our study shows that, even in an egal-

itarian developed country, social inequality has not been

completely banished from public health. The results show

that, even in the incidence of and survival from cancer, socio-

economic position plays a role. Our results are relevant for all

societies in which health interventions are established in

accordance with the goal of equal access to public services.

The results show that consideration must be given to how

inequalities in cancer incidence and survival can be ad-

dressed in public health interventions and indicate that fur-

ther research into the underlying mechanisms is required.

Conﬂict of interest statement

None declared.

Acknowledgement

The study was funded by a grant from the Danish Cancer

Society.

R E F E R E N C E S

1. Kogevinas M, Pearce N, Susser M, Boffeta P, editors.

Social

inequalities and cancer.

Lyon: International Agency for

Research on Cancer; 1997.

2. Dalton SO, During M, Ross L, et al. The relation between

socioeconomic and demographic factors and tumour stage in

women diagnosed with breast cancer in Denmark, 1983–1999.

Br J Cancer

2006;95:653–9.

3. Dalton SO, Ross L, During M, et al. Inﬂuence of socioeconomic

factors on survival after breast cancer—a nationwide cohort

study of women diagnosed with breast cancer in Denmark

1983–1999.

Int J Cancer

2007;121:2524–31.

4. Frederiksen BL, Osler M, Harling H, Jorgensen T. Social

inequalities in stage at diagnosis of rectal but not in colonic

cancer: a nationwide study.

Br J Cancer

2008;98:668–73.

5. Dalton SO, Steding-Jessen M, Gislum M, et al. Social

inequality and incidence of and survival from cancer in a

population-based study in Denmark, 1994–2003: Background,

aims, material and methods.

Eur J Cancer

2008;44:1938–49.

6. Boyle P, Parkin DM. Statistical methods for registries. In:

Jensen OM, Parkin DM, MacLennan R, Muir CS, Skeet RG,

editors.

Cancer registration: principles and

methods.

Lyon: International Agency for Research on Cancer;

1991. p. 126–58.

7. Andersen ZJ, Lassen CF, Clemmensen IH. Social inequality

and incidence of and survival from cancers of the mouth,

EUROPEAN JOURNAL OF CANCER

4 4 ( 2 0 0 8 ) 2 0 7 4

–

2 0 8 5

2085

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

pharynx and larynx in a population-based study in Denmark,

1994–2003.

Eur J Cancer

2008;44:1950–61.

Baastrup R, Sørensen M, Hansen J, et al. Social inequality and

incidence of and survival from cancers of the oesophagus,

stomach and pancreas in a population-based study in

Denmark, 1994–2003.

Eur J Cancer

2008;44:1962–77.

Egebjerg R, Halkjær J, Rottman N, et al. Social inequality and

incidence of and survival from cancers of the colon and

rectum in a population-based study in Denmark, 1994–2003.

Eur J Cancer

2008;44:1978–88.

Dalton SO, Steding-Jessen M, Engholm G, et al. Social

inequality and incidence of and survival from lung cancer in a

population-based study in Denmark, 1994–2003.

Eur J Cancer

2008;44:1989–95.

Carlsen K, Høybye MT, Dalton SO, et al. Social inequality and

incidence of and survival from breast cancer in a population-

based study in Denmark, 1994–2003.

Eur J Cancer

2008;44:1996–2002.

Jensen KE, Hannibal CG, Nielsen A, et al. Social inequality

and incidence of and survival from cancer of the female

genital organs in a population-based study in Denmark,

1994–2003.

Eur J Cancer

2008;44:2003–17.

Marsa K, Johnsen NF, Bidstrupp PE, et al. Social inequality

and incidence of and survival from male genital cancer in a

population-based study in Denmark, 1994–2003.

Eur J Cancer

2008;44:2018–29.

Eriksen KT, Petersen A, Poulsen AH, et al. Social inequality

and incidence of and survival from cancers of the kidney and

urinary bladder in a population-based study in Denmark,

1994–2003.

Eur J Cancer

2008;44:2030–42.

Birch-Johansen F, Hvilsom G, Kjær T, et al. Social inequality

and incidence of and survival from malignant melanoma in a

population-based study in Denmark, 1994–2003.

Eur J Cancer

2008;44:2043–9.

Schmidt LS, Nielsen H, Schmiedel S, et al. Social inequality

and incidence of and survival from tumours of the central

nervous system in a population-based study in Denmark,

1994–2003.

Eur J Cancer

2008;44:2050–7.

Roswall N, Olsen A, Christensen J, et al. Social inequality

and incidence of and survival from Hodgkin lymphoma,

non-Hodgkin lymphoma and leukaemia in a population-

based study in Denmark, 1994–2003.

Eur J Cancer

2008;44:2058–73.

Storm HH, Michelsen E, Clemmensen IH, Pihl J. The Danish

Cancer Registry: history, content, quality and use.

Dan Med

Bull

1997;44:535–9.

Dreyer L, Winther JF, Pukkala E, Andersen A. Tobacco

smoking.

APMIS

1997;105:9–47.

Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey SG.

Indicators of socioeconomic position (part 1).

J Epidemiol

Commun Health

2006;60:7–12.

Laaksonen M, Prattala R, Helasoja V, Uutela A, Lahelma E.

Income and health behaviours. Evidence from monitoring

surveys among Finnish adults.

J Epidemiol Commun Health

2003;57:711–7.

Huisman M, Kunst AE, Mackenbach JP. Educational

inequalities in smoking among men and women aged 16

years and older in 11 European countries.

Tob Control

2005;14:106–13.

Harper S, Lynch J. Trends in socioeconomic inequalities in

adult health behaviors among US states, 1990–2004.

Public

Health Rep

2007;122:177–89.

Glen P, Simpson MF, Donnelly L, Leonard S, MacDonald A.

Thirty-day mortality from colorectal cancer surgery within a

deprived population.

Colorect Dis

2005;7:193–5.

25. van Vliet EPM, Eijkemans MJC, Steyerberg EW, et al. The role

of socio-economic status in the decision making on diagnosis

and treatment of oesophageal cancer in the Netherlands.

Br J

Cancer

2006;95:1180–5.

26. Griggs JJ, Culakova E, Sorbero MES, et al. Effect of patient

socioeconomic status and body mass index on the quality of

breast cancer adjuvant chemotherapy.

J Clin Oncol

2007;25:277–84.

27. Griggs JJ, Culakova E, Sorbero MES, et al. Social and racial

differences in selection of breast cancer adjuvant

chemotherapy regimens.

J Clin Oncol

2007;25:2522–7.

28. Hansen RP, Olesen F, Sorensen HT, Sokolowski I, Sondergaard

J. Socioeconomic patient characteristics predict delay in

cancer diagnosis: a Danish cohort study.

BMC Health Serv Res

2008;8:49.

29. Neal RD, Allgar VL. Sociodemographic factors and delays in

the diagnosis of six cancers: analysis of data from the

‘National Survey of NHS patients: Cancer’.

Br J Cancer

2005;92:1971–5.

30. Lund L, Borre M, Jacobsen J, Sorensen HT, Norgaard M. Impact

of comorbidity on survival of Danish prostate cancer patients,

1995–2006: a population-based cohort study. Urology 2008 Mar

13 [e-pub ahead of print].

31. Tetsche MS, Dethlefsen C, Pedersen L, Sorensen HT, Norgaard

M. The impact of comorbidity and stage on ovarian cancer

mortality: a nationwide Danish cohort study.

BMC Cancer

2008;8:31.

32. Tetsche MS, Norgaard M, Jacobsen J, Wogelius P, Sorensen HT.

Comorbidity and ovarian cancer survival in Denmark, 1995–

2005: a population-based cohort study.

Int J Gynecol Cancer

2008;18:421–7.

33. Janssen-Heijnen MLG, Houterman S, Lemmens VEPP,

Louwman MWJ, Maas HAAM, Coebergh JWW. Prognostic

impact of increasing age and comorbidity in cancer patients: a

population-based approach.

Crit Rev Oncol Hematol

2005;55:231–40.

34. van Spronsen DJ, Janssen-Heijnen MLG, Lemmens VEPP,

Peters WG, Coebergh JWW. Independent prognostic effect of

comorbidity in lymphoma patients: results of the population-

based Eindhoven Cancer Registry.

Eur J Cancer

2005;41:1051–7.

35. Houterman S, Janssen-Heijnen MLG, Hendrikx AJM, van den

Berg HA, Coebergh JWW. Impact of comorbidity on treatment

and prognosis of prostate cancer patients: a population-based

study.

Crit Rev Oncol Hematol

2006;58:60–7.

36. Bronnum-Hansen H, Baadsgaard M. Increasing social

inequality in life expectancy in Denmark.

Eur J Public Health

2007;17:585–6.

37. Mackenbach JP, Stirbu I, Roskam A-JR, et al. Socioeconomic

inequalities in health in 22 countries.

New Engl J Med

2008;358:2468–81.

38. Sant M, Aareleid T, Berrino F, et al. EUROCARE-3: survival of

cancer patients diagnosed 1990–94—results and commentary.

Ann Oncol

2003;14(Suppl. 5):v61–118.

39. Coleman MP, Gatta G, Verdecchia A, et al. EUROCARE-3

summary: cancer survival in Europe at the end of the 20th

century.

Ann Oncol

2003;14(Suppl. 5):v128–49.

40. Berrino F, De Angelis R, Sant M, et al. Survival for eight major

cancers and all cancers combined for European adults

diagnosed in 1995–99: results of the EUROCARE-4 study.

Lancet

Oncol

2007;8(9):773–83.

41. Tretli S, Engeland A, Haldorsen T, et al. Prostate cancer—look

to Denmark?

J Natl Cancer Inst

1996;88:128.

42. Kvale R, Auvinen A, Adami HO, et al. Interpreting trends in

prostate cancer incidence and mortality in the ﬁve Nordic

countries.

J Natl Cancer Inst

2007;99:1881–7.